rick/LEDMatrix
1
0
Fork 0
mirror of https://github.com/ChuckBuilds/LEDMatrix.git synced 2026-10-04 14:25:08 +00:00
Code Issues Packages Projects Releases Wiki Activity
1,979 Commits 10 Branches 20 Tags
a6b3384032f90337906dd6c969d6c6bf6b62f7e7
Commit Graph
64 Commits
Author SHA1 Message Date
ChuckandClaude Opus 5 d51f7ada14 chore(scroll): drop the dead sub-pixel path, and two dev-tooling papercuts (#570) 
Three independent changes, none of which alter runtime behaviour.

1. Remove ScrollHelper._get_visible_portion_subpixel and
   _interpolate_subpixel (162 lines). get_visible_portion dispatches only to
   _blend_visible_portion, so _get_visible_portion_subpixel had no caller, and
   _interpolate_subpixel was reachable only from inside it -- a closed island.
   _blend_visible_portion's own docstring already records that the scipy path
   it replaced was dead; the replacement landed but the corpse stayed.

2. scripts/check_plugin.py: also search ../ledmatrix-plugins/plugins. The
   scoreboards live in the sibling checkout, so --all silently skipped every
   one of them and only --plugin-dir reached them.

3. .gitignore: ignore config/.config_secrets.json.tmp.*, which the suite
   leaves behind several of per run.

Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
 2026-09-14 09:42:50 -04:00
ChuckandClaude Opus 5 92ac231138 fix(fonts): load 4x6 on its pixel grid, from any working directory (#565) 
* fix(fonts): load 4x6 on its pixel grid, from any working directory

`extra_small_font` loaded 4x6-font.ttf at 6, off the face's 7px grid.
Under `draw.fontmode = "1"` the mono rasteriser thresholds each glyph at
50% coverage, so every glyph lost its fourth column and deformed:
christmas-countdown rendered "UNTIL" as "VM1JL". The advance is 5px at
both sizes, so snapping to 7 reflows nothing.

- Sizes in DisplayManager._load_fonts go through crisp_size() instead of
  literals. crisp_size / FONT_PIXEL_GRID / FONT_NAME_ALIASES move to
  src/common/font_layout.py; sports_card re-exports them.
- Mirror the fix in VisualTestDisplayManager, the harness's fork of
  _load_fonts. Without it every golden is blessed at the old size.
- Resolve bundled font paths against the install root, not the cwd.
  FontManager._resolve_asset_path now delegates to
  font_layout.resolve_asset_path (kept by name; plugins probe for it).
- The startup banner's middle rung snaps to 7; the 5 rung stays off-grid
  on purpose (the only size that fits a dotted quad on 64px).
- loading.py reads all plugin JSON as UTF-8 (cp1252 on Windows aborted
  check_plugin.py on a 0x9d byte).
- check_plugin.py reports in ASCII and never dies on an unencodable char.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(fonts): resolve relative asset paths from the install root, not the cwd

resolve_asset_path checked os.path.exists(relative_path) unconditionally,
so a relative asset path was still resolved against the process cwd first
-- exactly the dependency this module exists to remove. An unrelated
working directory that happens to contain assets/fonts/4x6-font.ttf (a
stale checkout, a copied assets folder, another project) would shadow the
real bundled font instead of the install root ever being consulted.

Only an absolute path is now returned as-is; a relative path always
resolves against _INSTALL_ROOT first, matching the docstring's stated
contract. FontManager._resolve_asset_path delegates to this function, so
it's covered by the same fix.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
 2026-09-13 10:47:01 -04:00
ChuckandClaude Opus 5 59997594ac test: fix the emulator port collision behind the intermittent suite failures (#562) 
* chore: stop tests and rigs writing to shared paths

Two shared-state problems, both of which show up as a permanently dirty
checkout or an unreproducible test failure.

test_display_dirty_tracking.py builds a real DisplayManager, whose
_snapshot_path defaults to the fixed /tmp/led_matrix_preview.png that the web
UI reads. Every pytest process on the machine shares that one file, so two
concurrent runs -- CI shards, a second worktree, an agent running the suite
alongside -- overwrite each other's snapshot and the mtime assertions stop
meaning anything. The module fixture now points it at a session-unique temp
path; the individual tests that care still override it further.

To be clear about what this does and does not fix: this is a real shared-path
hazard, but it is NOT the cause of the intermittent 15-test failure in that
module. That turned out to be the emulator's fixed TCP port, fixed in the
follow-up commit. This change stands on its own merits.

web_interface/app.py writes data/plugin_operations.json, data/plugin_state.json
and data/operation_history.json as the web interface runs, into a directory
that ships tracked (data/.gitkeep) and was otherwise unignored. So every rig
that ever opened the web UI -- and every test run that constructs the app --
left three untracked files behind and a permanently dirty `git status`. Only
data/.gitkeep is tracked under data/, so the negation keeps it.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* test: stop the emulator binding a fixed port, so concurrent runs can't collide

This is the cause of the intermittent full-suite failures we have been chasing:
runs of identical code landing anywhere between 100 and 130 failures, while
every implicated test passed in isolation.

Six test modules set EMULATOR=true and build a real DisplayManager. The repo's
emulator_config.json selects the "browser" adapter, which binds TCP port 8888 to
serve the dev preview. That port is a machine-wide singleton, so a second pytest
process -- a CI shard, another worktree, an agent running the suite alongside --
loses the bind. RGBMatrix construction then raises, DisplayManager catches it and
falls back to `self.matrix = None`, and every test that subsequently touches the
matrix dies with

    AttributeError: 'NoneType' object has no attribute 'SwapOnVSync'

which names neither a port nor a socket, and points at the wrong file entirely.
Because test_display_dirty_tracking's fixture is module-scoped, all 15 of its
matrix-touching tests fail together or not at all -- the 15-test swing that made
the totals look random.

Demonstrated rather than assumed. Holding 0.0.0.0:8888 from a separate process
and running test_display_dirty_tracking.py:

    without this change    15 failed, 6 passed
    with this change       21 passed

The "raw" adapter renders in memory and binds nothing. Only display_adapter is
overridden, in a throwaway config written per pytest process; the repo's
emulator_config.json is untouched and `run.py -e` still opens the browser
preview on 8888. Nothing in the suite referenced the adapter, and the tests
wrap SwapOnVSync on the matrix object itself, so they are indifferent to what
sits underneath. allow_adapter_fallback is forced off -- falling back would
land us on the browser adapter and its fixed port, which is the whole problem.

CONFIG_PATH is a bare relative filename resolved against the CWD, so it is set
to an absolute path: the previous behaviour depended on where pytest was invoked
from, and silently wrote a default config into whatever directory that was.

Verified no regressions: full suite on this branch and with origin/main's
versions of the touched files, same machine, back to back -- 115 failed /
4347 passed on both sides, zero failures unique to either. That 115 is the
pre-existing Windows-environment baseline (POSIX file modes, fcntl, shell
scripts, Linux-only binaries); CI on Linux remains authoritative.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* chore: mark the shell entry points executable

Eleven scripts shipped as 100644, so `./scripts/install/configure_web_sudo.sh`
fails with "Permission denied" and only works if you know to prefix `bash`.
That one matters most: the web UI's own error hint, added in #560, tells users
to run exactly that path when a system action fails for want of passwordless
sudo, and following that instruction verbatim did not work.

All eleven carry a shebang and are invoked directly, never sourced. The two
sourced libraries -- lib_lowmem.sh and lib_systemd_render.sh -- are deliberately
left non-executable, which is what distinguishes a library from an entry point.

Mode bits only, no content: 11 files changed, 0 insertions, 0 deletions. Applied
with `git update-index --chmod=+x` because this checkout is on Windows, where
core.fileMode is off and the working-tree bit is not tracked.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
 2026-09-11 15:45:13 -04:00
ChuckandClaude Opus 5 f6367d63ae security: triage the CodeQL backlog — 129 alerts, three of them live (#561) 
* fix(web): escape quotes in every HTML escaper, not just & < >

The escapers are all `div.textContent = x; return div.innerHTML`. That
round-trip escapes &, < and > -- the only characters the HTML serializer
must escape in a text node -- and leaves quotes alone. Every widget then
interpolates the result into a quoted attribute value:

    value="${escapeHtml(v)}"   title="${escapeHtml(v)}"

so a value of `x" onmouseover="alert(1)` closes the attribute and adds an
event handler of its own. CodeQL reported this 83 times
(js/incomplete-html-attribute-sanitization) across the widget files.

It is one bug, not 83: the widgets each carry a standalone fallback that
did escape quotes, but they all prefer BaseWidget.escapeHtml when
window.BaseWidget exists -- which it always does in the shipped page -- so
the correct fallbacks were dead code and the incomplete shared one ran.
Fixed at each source instead of at the call sites.

app-shell.js already documented this exact gap in a comment and worked
around it by building DOM nodes by hand; that workaround stays (setting a
property cannot be got wrong), the comment is now accurate.

cache.html's delete button interpolated the cache key into
`onclick="deleteCacheFile('...')"`. Escaping cannot help there -- the
browser HTML-decodes the attribute before parsing it as JS, so `&#39;`
becomes a real `'` again -- so the key moves to a data-cache-key
attribute that the handler reads back.

url-input.js additionally wrote a value straight into an <a href> after
validating it against a schema-supplied protocol list, and that list
accepted any RFC 3986 scheme -- "javascript" included. Scriptable schemes
(javascript, data, vbscript, blob, filesystem) are now refused both when
the list is normalised and when a URL is checked against it, and the
render path routes its href through the same check instead of emitting
whatever was stored (js/xss-through-dom).

test/js/unit/test_html_escaping.js reads each escaper out of the shipped
file and runs it, so losing the quote handling again fails a test rather
than a scan.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(security): stop request-supplied names from reaching paths outside their base

Three of the py/path-injection alerts were live, not lint:

* GET /api/v3/plugins/<plugin_id>/static/<path:file_path> read any file
  whose resolved path *string-prefixed* the plugin directory. Flask's
  default converter forbids a slash but not dots, and
  get_plugin_directory('..') returned the parent of the plugins directory
  because it exists -- so every file under the project root then prefixed
  that directory, config/config_secrets.json included. The prefix check
  was also wrong on its own terms: with plugin dir "plugin-repos/foo",
  "../foo-evil/x" resolves to "plugin-repos/foo-evil/x", whose string does
  start with "plugin-repos/foo".

* POST /api/v3/plugins/of-the-day/json/delete interpolated the request
  body's file_id into f"{file_id}.json" and unlinked it, unvalidated. A
  file_id of "../../../../etc/something" deleted that file. This is the
  one finding in the batch that destroyed data rather than exposing it.

* POST /api/v3/cache/delete passed the body's key through
  CacheManager.clear_cache to DiskCache, which joined it as a filename and
  called os.remove. Same shape, same result. The guard goes in
  DiskCache.get_cache_path, the single choke point get/set/clear share, so
  every caller is covered rather than just this route. Real keys are the
  stems of files already flat in the cache directory -- that is how
  list_cache_files derives them -- so nothing legitimate is turned away.

The rest of the cluster (web_interface/app.py's asset route, the plugin
update handler, _get_plugin_version, the plugin-schema read in config.py)
was guarded in ways that held, but each had grown its own version of the
check. They now go through one helper, src/common/path_safety.py, which
returns the *sanitised value* rather than a verdict -- so a caller cannot
validate one string and open another, which is how the two real bugs
above were shaped.

Also: WiFiManager.connect_to_network took the SSID and password straight
from POST /api/v3/wifi/connect into nmcli's argv. There is no shell there,
so CodeQL's py/command-line-injection alert overstates the risk -- but
nmcli reads a leading "-" as an option, so an SSID of "--ask" asks nmcli
to run differently rather than to join a network. Both values are now
checked for shape (802.11's 32-octet SSID limit, WPA's 8-63 char
passphrase or 64-char hex key, no control characters, no leading dash)
before any subprocess runs.

test/test_path_traversal_guards.py asserts on the filesystem, not just
the status code: a handler that returns 403 and deletes the file anyway
would pass the weaker check. Twelve of its cases fail against the
unpatched code.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(web): refuse a plugin id that is not a plain name, don't truncate it

pages_v3 and scripts/dev_server.py ran request ids through
os.path.basename and carried on with what came out, so "../weather"
rendered the config form for "weather". Nothing escaped the plugins
directory -- the relative_to guards held -- but the handler answered a
request nobody made, and validating one string while the filesystem sees
another is the shape both live traversals earlier in this branch had.

Same treatment as the rest: safe_path_component rejects rather than
truncates, resolve_under returns the path it checked, and the call sites
use what those return. The three handlers that had hand-rolled
resolve-and-relative_to blocks lose about twenty lines to the shared one.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* docs(web): say what the plugin web_ui iframe actually is

The docstring claimed the fragment runs "in a sandboxed iframe". The
iframe in plugin_config.html carries no sandbox attribute, so the
fragment runs with the interface's own origin. That is fine -- the file
belongs to an installed plugin, and an installed plugin already runs
Python on the device, so the trust boundary is install rather than this
route -- but a comment promising containment that is not there is worse
than no comment. This is the context for the py/reflective-xss alert on
this handler.

Also drops the now-unused os/os.path imports.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(web): inline url-input's scheme guard at the previewLink.href sink

CodeQL flagged this line as a new high-severity js/xss-through-dom alert
on this PR even though it is already covered by SCRIPTABLE_SCHEMES: the
guard reached the sink through safeHref -> isValidUrl, two function calls
away, which its DOM-based-XSS sanitizer recognition does not trace.

Behavior is unchanged -- same scheme check, same SCRIPTABLE_SCHEMES list,
same allowedProtocols gate -- just inlined directly above the
previewLink.href assignment it guards, so the barrier is visible in the
same scope as the sink.

Added a regression test that runs the shipped onInput handler (not just
the extracted helpers) against a mocked DOM, so a future change that
reintroduces an unguarded previewLink.href assignment fails here.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(security): address CodeRabbit findings on the CodeQL triage PR

- src/wifi_manager.py: reject non-ASCII WPA-PSK passphrases before any
  credential-saving or connect flow runs. NetworkManager only accepts
  printable ASCII passphrases (or a 64-char hex key); a non-ASCII value
  was previously saved/attempted before nmcli itself rejected it.

- web_interface/blueprints/api_v3/config.py: fail closed when the
  plugin config schema path can't be resolved under the plugins
  directory (e.g. a symlinked plugin dir). Previously this fell
  through with secret_fields left empty, so submitted credentials for
  that plugin were saved as ordinary, unencrypted configuration.

- web_interface/static/v3/js/widgets/plugin-file-manager.js: stop
  splicing the JSON day/column key into an inline oninput="..." handler
  string. escHtml() escapes quotes for a normal HTML attribute, but the
  browser HTML-decodes the attribute before running it as script, which
  undoes that escaping and lets a crafted column name (e.g. from an
  uploaded JSON file) break out of the JS string and execute. Cell
  edits now travel through data-day/data-col attributes read by one
  delegated 'input' listener instead.

  While in this file: fixed 6 pre-existing missing-')' typos on
  multi-line safeSetHTML(...) calls (already flagged by Biome in this
  PR's own CodeRabbit run as syntax errors blocking its lint pass).
  These predate this PR (present on main too) but made the whole file
  fail to parse in any JS engine, which is a bigger problem than the
  XSS finding itself and directly touches the same lines.

Added/extended regression tests for each fix; full suites pass
(pytest: 4580 passed, 62 skipped; JS: 84 assertions).

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
 2026-09-11 15:32:03 -04:00
ChuckandClaude Opus 5 5137e86d16 feat(tools): MQTT bridge and Pixlet editor, ported onto the api_v3 split (#554) 
* feat(tools): manage the MQTT bridge and Pixlet editor from the Tools tab

PR #544's change, ported onto the api_v3 package split (#553). Identical
behaviour; only the placement of the new code differs.

The original added 508 lines to web_interface/blueprints/api_v3.py, which #553
deletes, so every hunk of it would conflict irreconcilably. Ported by AST:
26 new top-level items sorted to where the split puts each kind --

  __init__.py   2 imports, 11 constants, 7 helpers
  starlark.py   4 routes  (/starlark/editor/{apps,status,start,stop})
  misc.py       2 routes  (/integrations/mqtt-bridge{,/config})

Everything outside api_v3.py -- the Tools partial, the installer scripts, the
JS tests -- applied unchanged.

Routes: 111 from the split plus these 6 = 117, and the url-map snapshot is
regenerated to match, which is exactly what test_api_v3_url_map.py is designed
to make you do when routes are added.

Full Python suite: 4,278 passed, 68 skipped, 0 failed. The JS tests this PR
ships could not be run here -- node is not installed on this machine -- so
test/js/dom/test_tools_sections.js is unverified.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_014RRtqXDCnvnY6EQwhT5CV9

* fix(starlark): don't crash the pixlet editor's start/stop routes, and honor an operator-set PIXLET_EDITOR_HOST

The AST-based port of #544 onto the api_v3 package split dropped `time`
from starlark.py's import list. start_pixlet_editor() and
stop_pixlet_editor() both call time.time()/time.sleep() directly, so
every start (NameError building `state['started_at']`) and every stop
that has to wait out the EXIT trap crashed with a 500. No test caught
it because the route's own tests mock subprocess.Popen but never
actually invoked it before now.

Also carries over #544's later fix that this port branched before:
env['PIXLET_EDITOR_HOST'] = '0.0.0.0' unconditionally overrode an
operator who had already pinned PIXLET_EDITOR_HOST to loopback,
forcing the unauthenticated `pixlet serve` process onto the LAN
regardless (CodeQL CWE-1188). Switched to env.setdefault(...), same as
api_v3.starlark.py's siblings already do for _pkg-owned names.

Both fixes route the shared _pkg.time reference the rest of the
package's route modules already use for anything a test might need to
patch, rather than a bare `import time` local to this file.

Ported the existing regression test from #544
(TestPixletEditorHostDefaultsButDoesNotOverride) onto this branch's
module layout (web_interface.blueprints.api_v3.starlark instead of the
old monolithic api_v3 module), which is what caught the NameError.

Full suite: 4330 passed, 62 skipped, 2 failed -- identical on this
branch and on origin/main (missing tzdata package breaks two
timezone-alias tests in test_onboarding_checklist.py, unrelated to
this change).

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(api-v3): clear the six lint errors this rebase introduced

All six were introduced by rebasing this branch onto the merged blueprint
split, not by the split itself. Confirmed by diffing pyflakes output against
main with line numbers normalised -- everything else it reports is present on
main too and is the package's deliberate re-export pattern.

starlark.py used _STARLARK_APPS_DIR three times without importing it (F821).
The rebase resolved an import-list conflict as a union of both sides, and that
symbol was on neither side of the conflict hunk, so it was silently lost. It is
defined in __init__.py and is now imported like its neighbours. This was the
only one of the six that would fail at runtime rather than merely lint.

__init__.py imported contextlib twice (F811): the cherry-pick added one next to
the existing import. Removed the duplicate; the original at line 19 is used.

__init__.py imported signal purely to re-export it to starlark.py, so pyflakes
saw it as unused (F401). signal is stdlib and does not need routing through the
blueprint package, so starlark.py imports it directly and __init__.py no longer
does. contextlib stays re-exported because this module genuinely uses it.

_read_mqtt_bridge_config()'s local `config` shadowed the `config` submodule
this module imports at the bottom for its route side effects (F811). Renamed to
`settings`, with a comment saying why, since the name is otherwise the obvious
one to reach for.

Verified: pyflakes now reports nothing on this branch that main does not, the
package imports, all nine route modules load, and 117 routes register, matching
the pinned URL-map snapshot.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(api-v3): reject MQTT bridge bodies the endpoint cannot apply

Two CodeRabbit findings on the bridge settings endpoint, both of which returned
200 while doing something other than what the caller asked.

`request.get_json(silent=True) or {}` turned a missing or unparseable body --
and the JSON literals null, [] and false -- into an empty dict, which then
satisfied the isinstance(data, dict) guard on the very next line. The guard was
there to reject exactly those bodies. Dropping the `or {}` lets None fail it.

The same `or {}` on /errors/clear is left alone: its docstring documents the
body as optional, so an absent body legitimately means "use the defaults". The
difference is that saving settings has nothing sensible to do with no body.

`if data.get('clear_password'):` accepted any truthy value, and the string
"false" is truthy in Python -- so a client echoing the field back as a string
wiped a password it meant to keep. Now coerced through the package's existing
_coerce_to_bool, which already maps 'true'/'on'/'1'/'yes' and nothing else.

test_mqtt_bridge_config_endpoint.py covers both: five unusable body shapes plus
a missing body, and clear_password across truthy and falsy spellings. Verified
against the unfixed code -- reverting the body guard fails 5, reverting the
coercion fails 3.

Not changed here: CodeRabbit also asks this endpoint to reject MQTT credentials
when TLS is off (CWE-319). That is a policy decision about the feature rather
than a defect -- unencrypted MQTT on a trusted LAN is common and often
deliberate -- so it is raised on the PR for a maintainer call instead.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix: work through the remaining review findings on the editor and bridge

allow_insecure_mqtt (CWE-319, requested): a password with TLS disabled crosses
the network in cleartext. Refused now rather than merely warned about -- but
refused, not forbidden, because unencrypted MQTT on a trusted LAN is a normal
deliberate setup. allow_insecure_mqtt is the explicit acknowledgement, defaults
false, and is coerced like the other booleans so the string "false" cannot
switch the guard off.

starlark.py:796 -- the supported service runs Flask threaded, so two start
requests could each see running=False, each launch an editor, and the second
state write replace the first PID, orphaning a process that holds the display
down with nothing recording it. The check-launch-write sequence now takes a
module-level lock.

starlark.py:848 -- if the state write failed the route returned success with an
editor running and no PID recorded: status and stop both reported no session
while the display stayed down until the timeout expired. It now terminates the
process group and returns an error.

starlark.py:890 -- SIGKILL gives the script's EXIT trap no chance to run, so
nothing hands the display back, yet the response said "the display is
restarting". After an escalation the display is now restarted explicitly, and a
failure to do so returns an error naming the manual step instead of a success.

pixlet_config_editor.sh:184 -- find_pixlet supports Darwin but macOS ships no
timeout(1); GNU coreutils installs it as gtimeout. Resolved up front so the
failure lands before the display is stopped rather than after.

pixlet_config_editor.sh:154 -- wildcard, loopback and an explicit interface
address are three cases, not two. Collapsing the last two printed a URL saying
"localhost" whenever PIXLET_EDITOR_HOST named a LAN address.

tools.html:1254 -- escHtml does not encode single quotes, and the app id was
interpolated into an inline onclick="startPixletEditor('...')", so a directory
containing an apostrophe could break out of the JS string and run script. The
handler binds with addEventListener and reads the id from dataset, where it is
only ever parsed as an HTML attribute.

Tests: test_mqtt_bridge_config_endpoint.py grows to 23 cases covering the opt-in
in both directions. The tools DOM suite gains three guards asserting the edit
buttons carry no inline onclick and pass the id via dataset -- those need jsdom
and did not run here, so CI verifies them.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(api-v3): log the traceback on the editor state-write failure

The 848 fix answers 500 when the session state cannot be written, and logged
that at error level -- but without exc_info, so the traceback never reached the
log. test_web_error_detail.py guards exactly this: a handler returning 5xx must
write an error-level record *with* the traceback and return the sanitized
detail, because checking that merely something was logged is too weak.

Caught by Core unit tests on the previous commit, not locally: the guard parses
every module under web_interface/blueprints/api_v3 as one source, so it only
fires once the whole package is read together.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
 2026-09-11 11:32:12 -04:00
ChuckandClaude Opus 5 12f3790994 fix(install): render the systemd units from their templates, not from heredocs (#547) 
* fix(install): render the systemd units from their templates, not from heredocs

The installers carried their own inline copies of units that also exist as
templates under systemd/, and the copies drifted.

install_service.sh renders ledmatrix.service from the template correctly, then
wrote ledmatrix-web.service from a heredoc that predated it -- missing
Wants=network-online.target, RestartSec=10, SyslogIdentifier, CacheDirectory,
CacheDirectoryMode and Environment=USE_THREADING=1. install_web_service.sh had
a third copy, and install_wifi_monitor.sh a fourth, that one already differing
from its template (syslog where the template says journal).

startup_validator.py compares the installed unit against the template, so a
rig installed this way warned on every boot -- and the remedy the warning
names, "re-run scripts/install/install_service.sh", reinstalled the same stale
copy. The warning could never clear. Reproduced on a live rig running exactly
that unit.

All three installers now render systemd/*.service through the same placeholder
substitution. The template gains a __USER__ placeholder rather than hardcoding
User=root, because the web interface runs as whoever installed it.

That last point was a second, independent cause of a permanent warning: the
validator substituted a fixed "root", so any non-root install reported drift
forever. It now reads User= from the installed unit -- an install-time
decision, not something the template dictates -- and compares everything else
strictly. first_time_install.sh already reads the installed User= the same way.

Tests cover a non-root web unit not warning, a genuinely changed directive in
that unit still warning, the User= fallback, and a grep-based guard that no
installer under scripts/install/ contains an inline unit body. That guard is
what found the install_wifi_monitor.sh copy.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_014RRtqXDCnvnY6EQwhT5CV9

* fix(install): escape sed replacements, use mktemp, and make render failures fatal

Address CodeRabbit findings on install_service.sh, install_web_service.sh and
install_wifi_monitor.sh:

- Values interpolated into each script's sed expression (project root path,
  username) were not escaped, so a value containing &, \ or the | delimiter
  would corrupt the rendered systemd unit. Add a shared
  sed_escape_replacement() helper in the new scripts/install/lib_systemd_render.sh
  (sourced by all three scripts) and apply it to every sed replacement.
- install_service.sh rendered the main and web units to the predictable path
  /tmp/ledmatrix.service.tmp before installing them -- a symlink/TOCTOU race
  (CWE-377). Use mktemp for both, with a trap to clean up on exit.
- install_service.sh treated a missing template as a mere warning and then
  checked only whether a unit already existed at the destination before
  enabling/starting it, so a render failure could silently fall back to
  enabling a stale, previously-installed unit. Both unit blocks now exit
  non-zero on a missing template or a failed render.

Also rename the ambiguous loop variable `l` to `line` in
test/test_systemd_unit_drift.py (Ruff E741); ruff isn't wired into any CI
workflow in this repo today, so this isn't currently CI-blocking, but the
rename is trivial and correct regardless.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01S3bPMESe2TfrGvbs1ef9c5

* test(install): cover sed_escape_replacement against sed-special characters

CodeRabbit asked for regression coverage using a project path containing an
ampersand; the earlier commits on this branch already fixed the escaping,
mktemp usage, and enable/start-on-fatal-render-failure findings, and the
l->line rename was already applied -- this closes the one remaining gap.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
 2026-09-11 08:41:19 -04:00
ChuckandClaude Opus 5 2df273ecfc fix(web): default web_display_autostart to true, as the installer already does (#556) 
start_web_conditionally.py read the flag with
`config_data.get("web_display_autostart", False)`, so a config that simply
lacked the key got no web interface. Both config/config.template.json and
first_time_install.sh ship the key as true, so the code default contradicted
the shipped default in two places: absence means an older or hand-edited
config, not a request to stay down.

The failure mode was silent in the worst way. The "not starting" path exits 0,
so `systemctl status ledmatrix-web` reported the unit as successfully started
while nothing was listening on the port, and the only trace was one journal
line saying the flag was "false or not set" -- which reads as a deliberate
setting rather than a missing key.

Also start the web interface when config.json is missing or unparseable,
instead of exiting. The web interface is how a config gets created and
repaired, so a broken config is exactly when the user needs it most; leaving
it down means there is no way back in. Only an explicit false/off disables
autostart now, and the disabled message says "explicitly disabled" so the
journal distinguishes a real setting from a default.

Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
 2026-09-10 17:42:00 -04:00
ChuckandClaude Opus 5 f8e2e89edc refactor(sports): put the scoreboards on the shared scroll resolver (#542) 
* refactor(sports): put the scoreboards on the shared scroll resolver

Eight sports scoreboards -- afl, baseball, basketball, football, hockey,
lacrosse, nrl, soccer -- scrolled through this module's own pacing while the
other eleven scrolling plugins went through src/common/scroll_config. Two
implementations of the same job, and this one was on the losing side of every
difference.

It never called set_scrolling_state. Two consequences, both of which this
release's work was about:

- The frame hold is applied through that call, so a speed the crisp ladder
  could render in whole pixels still presented a new frame every refresh.
- Core only runs deferred updates while nothing is scrolling. Believing
  nothing was, it ran blocking work in the middle of these scrolls.

The default is non-crisp today: scroll_speed 50.0 with scroll_delay 0.01 is
50 px/s, which on a 100Hz panel is half a pixel per refresh. That cannot
render as motion -- it alternates 0px and 1px steps and judders at a 50Hz
beat, on every scoreboard, out of the box. Resolved through the ladder it
stays 50 px/s and holds each frame for two refreshes: same speed, whole-pixel
motion.

The stepping disagreement that used to justify a separate module is gone.
scroll_config avoided frame-based mode because it stepped on a wall clock at
1/scroll_delay with scroll_delay set to the frame period, so the decision sat
on its own threshold and flipped on sub-millisecond jitter. That branch now
accumulates elapsed time, identical arithmetic to the time-based one, so the
two differ only in the units the speed arrives in.

What is NOT shared, and must not be: the two modules read identically-named
keys with different meanings. Here scroll_speed is px/SECOND and scroll_delay
only converts to px/frame; in scroll_config scroll_speed is px per STEP, so
px/s is speed/delay. Passing this module's settings dict to the resolver turns
50 px/s into 5000, clamped to 500 -- a tenfold speed-up everywhere. So
_get_scroll_settings keeps sole ownership of reading sports config, including
the league merging, and hands the resolver a plain px/s. A test pins that
specific number, because it is the mistake the refactor invites.

MIN/MAX_PIXELS_PER_FRAME are gone; the resolver bounds speed and the helper
clamps FPS. _resolve_target_fps stays, re-purposed: under the old model that
key was the rate frames were presented at, so it is the faithful translation
into the refresh the ladder is computed against, used when no hardware
refresh is configured.

Speed changes for panels that are not 100Hz: 50 px/s becomes 60 at 60Hz
(+20%) and 48 at 120Hz (-4%). At 100Hz it is unchanged.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(display): drop the frame hold when a scroll times out, not just when it says so

set_scrolling_state(False) clears the hold. The other way a scroll ends is
is_currently_scrolling() deciding, after scroll_inactivity_threshold of
silence, that it is over -- which is what happens when the rotation moves on
mid-scroll or a plugin is torn down. That path cleared the flag and kept the
hold, so every later plugin, scrolling or static, was presented at refresh/N
by whoever scrolled last, until something called the explicit stop.

The method's own docstring already states the rule this breaks: the hold "must
not outlive the scroll that asked for it". The timeout was the exception it
did not cover.

Pre-existing, but reachable by three plugins before and eleven after the
sports scoreboards moved onto the shared resolver, so it belongs with that
change. The test ages the activity timestamp past the threshold rather than
sleeping.

Also adds scripts/sports_scroll_check.py. The sports scroll path is per-league
opt-in, so a rig showing static game cards never constructs a
SportsScrollDisplay and none of its pacing can be observed from a normal run
-- which is exactly what happened when this change was first put on hardware:
26 minutes, zero sports scroll lines. The script drives the path directly with
synthetic games and asserts the three things the resolver is meant to buy: the
speed lands on whole pixels, the hold is published, and it is released after.
It never starts or stops the display service, matching scroll_speeds.py, so a
crash here cannot leave the panel dark.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(scripts): refuse to grab the panel while the display service has it

The module docstring already said to stop ledmatrix first. Nothing enforced
it, and running the script against a live service is not a harmless mistake:
rpi-rgb-led-matrix configures GPIO directions and the hardware PWM inside
RGBMatrix(), and when the root check fails it calls exit() from C with no
cleanup. The service keeps rendering and swapping onto pins that have been
reconfigured underneath it, so the panel goes black while every diagnostic
says the display is healthy -- fresh framebuffer, every pixel lit, "RGB Matrix
initialized successfully", nothing in the log. A restart fixes it, once you
work out that is what happened.

Found the hard way: this is what took the panel down on the test rig, not the
change the script was written to verify.

--fallback skips the check, since it never opens the matrix. --force is there
for anyone who means it.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* chore(scripts): annotate the subprocess call the way this repo already does

Codacy fails a PR on one new issue, and bandit B404 fires on any subprocess
import. scripts/run_plugin_tests.py carries the same suppression with the same
justification -- list-form argv, no shell -- so this follows it rather than
inventing a second convention.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
 2026-09-09 16:13:04 -04:00
ChuckandClaude Opus 5 968b953a51 fix(display): pin one text layout engine, and give the 5x7 BDF face a size (#539) 
* fix(display): pin one text layout engine, and give the 5x7 face a size

Two ways a font could render differently on two machines running the same
code, both found while diagnosing four plugins whose golden images passed on
the machine that generated them and failed everywhere else.

**Layout engine.** `ImageFont.truetype` picks its engine at load time: Raqm
where the host Pillow was built with libraqm, Basic otherwise. The two round
fractional glyph advances differently. `PressStart2P-Regular.ttf` at 8px has
whole-pixel advances, so they agree — which is why most of the fleet matched
everywhere and hid this. `4x6-font.ttf` at 6px does not: glyph positions drift
cumulatively along a run, and the four plugins that draw body text in it
(geochron, of-the-day, christmas-countdown, ledmatrix-weather's almanac) are
exactly the four whose goldens travelled badly.

Every core font load now goes through `src/common/font_layout.load_truetype`,
which pins the Basic engine, so a render depends on the font file and the size
and nothing else. Basic gives up complex-script shaping and kerning pairs;
neither applies to bitmap-grid faces on an LED panel. Output is unchanged on a
host without libraqm.

**Zero font height.** `DisplayManager` built the 5x7 BDF face with
`freetype.Face(path)` and never called `set_char_size`, so `face.size.height`
stayed 0 and `get_font_height()` returned 0 for it — callers stacking rows by
`prev_y + prev_height + gap` drew two lines on top of each other. The
start-up line `Calendar font size: 0 pixels` has been printing the symptom all
along. `font_manager._load_bdf_font` already called `set_char_size`, so
whether measurement worked depended on which path loaded the face.

`DisplayManager` now sets it too, and `get_font_height()` falls back to the
strike the file declares rather than returning a zero line height.

Fixes ChuckBuilds/ledmatrix-plugins#397
Refs ChuckBuilds/ledmatrix-plugins#371, #375, #378, #391

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(display): give the startup banner a rung that fits a full address at 64px

CI caught what pinning the layout engine exposed rather than caused.
`_fitting_font` walks PressStart2P then 4x6 at 6px, and "255.255.255.255" --
the widest thing the startup banner ever shows -- measures 66px at 4x6/6px
against the 62 a 64x32 panel has to give. It used to squeak in only because
the measurement depended on which layout engine the host Pillow happened to
have; with the engine pinned it does not, so the rung the worst case actually
needs is now in the ladder instead of implied: 4x6 at 5px, which measures 51.

The fallback was wrong in the same place. When nothing in the ladder fit, it
returned `self.font` -- the *widest* option, and precisely how "Initializing"
came to run off the side of a 64px panel to begin with. It returns the
narrowest face that loaded now.

test/test_initializing_screen.py: 34 passed.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(display): name the exceptions the BDF strike read can raise

Codacy flagged the try/except/pass. It was already narrow in intent -- a
malformed strike table on the measurement path must degrade to "size unknown"
rather than take the display down -- but a bare `except Exception: pass` says
neither of those things and hides a genuinely broken font behind a silent 8px
fallback. It now catches what reading `available_sizes` can actually raise and
logs which face failed.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* chore: drop logo PNGs the render harness downloaded into the worktree

These are fetched at runtime by the logo cache; they are not source, and they
rode in on a `git add -A` while I was running check_plugin.py against this
branch. Nothing in the change needs them.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
 2026-09-08 09:40:00 -04:00
ChuckandClaude Opus 5 e23f1f45d3 feat(starlark,on-demand): the third-party fixes worth taking, plus a Home Assistant MQTT bridge (#538) 
* feat(starlark,on-demand): the third-party fixes worth taking, plus an MQTT bridge

Analysis of ant456/ledmatrix-fixes-repo, a third-party collection of
patches and services built while running this project on Starlark apps
under MQTT control. Its patches are whole-file copies taken against an
older tree, so applying them as written would revert #523's frame
pacing, #534's display() bool returns and the GitHub token masking in
plugins_manager.js. Three of its claimed fixes are already in main, and
its api_v3 Starlark routes are #535's. What follows is the rest --
verified against current code, and reimplemented where the patch's
approach did not hold up.

**On-demand display.** `pinned` reached the controller from the API, was
stored on it and republished in the status payload, but never narrowed
the rotation -- a pinned request still cycled every mode its plugin
owns. Right for a sports plugin, whose modes are views of one subject;
wrong for a plugin whose modes are unrelated, which is every Starlark
app. Now honoured, and it survives a restart.

Restarting while on-demand was active loaded *only* the on-demand
plugin, so normal rotation had nothing to return to for the life of the
process -- and a restart mid-session is routine, since that is how an
update is applied. The panel came back cycling one plugin's modes with
no way out but clearing the cache by hand. Every enabled plugin loads
now; on-demand still resumes on its saved mode.

Stop requests are exempt from the duplicate guards on purpose, so that a
second click stops a mode a race left running -- which means consuming
the mailbox is the only thing that ends one. It was never consumed, so
the same stop was re-read and re-processed on every poll, forever. Both
paths now share one compare-before-delete helper.

**Starlark rendering.** `extract_schema` parsed the source with a regex,
which can only see option lists written out literally: an app whose
dropdown is filled from a live API call inside `get_schema()` came back
empty, and the config form offered nothing to pick. Now runs `pixlet
schema`, which executes the app, and falls back to the parser when
Pixlet is absent, too old for the subcommand, or the app fails to run.
The third-party patch replaced the parser outright and hardcoded
/usr/local/bin/pixlet; this keeps the fallback and the binary search.

A `|` in a config value was dropped by a shell-metacharacter filter,
though the command is a list with no shell involved -- and apps do use
it as a separator inside one value. The key went missing silently and
the app rendered its own "not configured" screen with nothing to say
why. And a 0-byte render was reported as success: Pixlet exits 0 and
writes nothing when an app has no content, which read downstream as a
working app drawing a black panel.

**Starlark display.** `display()` ignored the mode it was called with,
so a specific app could not be addressed. It now accepts `display_mode`
-- which is the whole mechanism, since the controller inspects the
signature before passing it. Found while there: `_select_next_app` ran
only while `current_app` was unset, so with several apps installed the
first was picked once and shown forever while the rest were rendered on
schedule and never displayed. And `enable_scrolling` was missing, so
multi-frame apps were called once per rotation slot and never advanced
past frame one.

**GET /api/v3/display/modes.** Every mode that can be requested
on-demand, with the plugin that owns it. Nothing exposed this, so
anything driving the display from outside the web UI read each plugin's
manifest.json off disk and reimplemented PluginManager's fallbacks. It
also triggers discovery, which is otherwise lazy and normally happens
because a person opened the dashboard.

**integrations/mqtt_bridge.** Home Assistant control over MQTT
Discovery: a mode select, a stop button, power, brightness. Rewritten
against the API rather than the filesystem, so it needs no read access
to config.json and cannot drift from the web UI. paho-mqtt 2.x
VERSION2, TLS, an availability topic that is also the last will, and
secrets from the environment.

**Two opt-in extras.** A DNS single-request unit, for glibc's parallel
A/AAAA lookup stalling ~5s per name on routers that answer only the A
query -- which makes any plugin calling an external API slow and
Starlark apps, which have a render timeout, fail outright. And a Pixlet
config editor: a script you run and Ctrl+C rather than the third-party
version's always-on unauthenticated Flask service, since it stops the
display for the length of a session. Neither is installed by default.

Long Starlark app names now wrap instead of overflowing their card.

115 new tests across 5 files. Also unblocked
test_starlark_display_contract.py, which was silently skipping wherever
fcntl is absent. Whole suite: no new failures against main.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(mqtt_bridge): the five issues Codacy flagged on this branch

All in the new bridge, all real:

  * requests floor was 2.31.0, which carries CVE-2024-35195,
    CVE-2024-47081 and CVE-2026-25645. Raised to >=2.33.0,<3.0.0, which
    is what the project's own requirements.txt already pins.
  * `import time` was never used.
  * `"mqtt_password": None` in DEFAULTS read as a hardcoded credential.
    It is the "no password configured" default; marked nosec B105, the
    convention used elsewhere in the repo.

Also dropped an unused `build_app` from the display-modes test imports.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix: the review findings on this PR

Nine of CodeRabbit's ten, plus the CodeQL alert. The tenth is wrong and
is answered below.

**One bad config section blanked the whole mode list.**
`/display/modes` read `full_config.get(plugin_id, {}).get('enabled')`,
so a non-dict under a plugin id -- a shape DisplayController already
guards, so it happens -- raised AttributeError mid-loop and answered 500
with no modes at all. Every MQTT bridge entity is built from that list.
Now skipped with a warning.

**The DNS scripts reported success they had not earned.** Three separate
paths: `resolvconf -u` failing was swallowed by `|| true`; the
systemd-resolved branch exited 0 without applying anything, so the
oneshot unit recorded success while the workaround was inactive; and the
installer's `|| echo` turned a failed start into "installation
complete." with exit 0. All three now fail loudly. `single-request` is a
glibc resolv.conf option with no resolved.conf equivalent, so on those
hosts the honest answer is that it cannot be applied.

A NetworkManager-generated resolv.conf is regenerated on connection
changes, not only at boot, and the unit is oneshot with RemainAfterExit
-- so the option can vanish mid-boot with nothing to put it back. Now
detected and stated plainly rather than implied to be permanent.

**`Before=` does not order a manual restart.** It only orders units
already in the same transaction, so `systemctl restart ledmatrix` could
bypass the fix. install_dns_fix.sh now writes a ledmatrix.service
drop-in with Wants= and After=. Wants=, not Requires=: a DNS workaround
failing should not stop the display.

**The Pixlet editor's `--lan` is gone.** `pixlet serve` has no
authentication, and a printed warning is not access control. Loopback
only, with the SSH port-forward in the header where the flag used to be
documented -- SSH does the authenticating and nothing is left listening.

**The MQTT example config now defaults to TLS** on 8883. The installer
copies it verbatim, and without TLS the broker password and every
command cross the network in cleartext. A plaintext broker is still
supported and documented, and the bridge warns once at startup when a
password is configured without TLS.

**Not taken: "the upstream Pixlet CLI has no `schema` subcommand."**
Upstream tidbyt/pixlet has none, but `scripts/download_pixlet.sh`
installs `tronbyt/pixlet`, whose `cmd/schema.go` is
`schema [PATH]` -> JSON on stdout, built on
`runtime.NewAppletFromPath`, so it does execute `get_schema()`. That is
exactly what extract_schema_via_pixlet calls. A binary without the
subcommand exits non-zero and falls back to the source parser, which is
already covered by a test.

**CodeQL stack-trace exposure: not taken either.** I removed `details`
first and that broke
test_web_error_detail.py::test_no_api_v3_handler_discards_its_exception,
which enforces `describe_exception` across all ~75 handlers -- written
because a device with failing storage answered "see logs for details"
from the log viewer itself. describe_exception redacts credentials; the
trade-off is the project's and is already made. Restored, with the
reasoning in a comment.

11 new tests. Whole suite: no new failures against main, 4127 passed.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
 2026-09-08 08:34:52 -04:00
ChuckandClaude Opus 5 c9289e3a1d fix(store): update_plugin silently did nothing for four installed plugins (#536) 
install_plugin() deliberately renames a plugin's directory to the MANIFEST id
when it differs from the REGISTRY id, so registry `stocks` lands in
`ledmatrix-stocks/`. Every lookup in _find_plugin_path() is by directory name,
so update_plugin("stocks") found nothing, logged "Plugin not installed", and
returned False.

Nothing surfaced that to the user. Clicking update in the web UI was a no-op
with no error, and the plugin stayed on a stale version indefinitely. Four
installed plugins hit this on a real device -- leaderboard, music, stocks and
weather -- found because a scripted update of eleven plugins failed on exactly
those four.

Adds a manifest-id scan as the LAST step of the resolution chain, so the two
documented lookups above it (configured dir, then the sibling plugins/
fallback) keep their exact meaning and ordering. That ordering is pinned by
test_discovery_path_contract.py, which characterises the divergence between
the three resolvers on purpose; this extends the chain rather than reordering
it. Directories renamed aside with '.standalone-backup-' during an install or
rollback are skipped, since matching one would report a half-finished install
as a live plugin.

Also adds scripts/audit_render_path.py, which walks the call graph from
display() and reports blocking calls reachable from it. display() runs on the
render thread, so anything slow there stalls the panel; on a vsync-paced loop
a single 15ms call drops a frame and a network round trip freezes the marquee.
Two instances were already found the slow way, by reading frame-time
histograms -- odds-ticker reading the scoreboard cache per frame, and
soccer-scoreboard timing out inside update(). The audit finds that shape in
the source instead. It is a heuristic and says so: a hit behind an interval
check may be fine.

It currently flags 23 calls across six plugins. The clearest is
ledmatrix-music, whose display() falls back to an inline
requests.get(timeout=5) when album art has not been prefetched -- a deliberate
"show the art rather than go blank" tradeoff by its author, but up to five
seconds of frozen panel. Reported, not changed; that is its owner's call.

185 store tests pass. Three of the six new tests fail without the fix.

Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
 2026-09-07 15:55:46 -04:00
ChuckandClaude Opus 5 d12323e7f1 perf(scroll): pace frames to the panel — 44→100 fps, stalls 14% → 0.02% (#523) 
* perf(scroll): pace frames to the panel, not to a fixed sleep

Scrolling ran at 44-46 fps on a 2x128x64 chain and 14-17% of frames took
41-53ms, which reads as judder. Four independent causes, each measured on
the hardware; details and the diagnostic recipe are in
docs/SCROLL_PERFORMANCE.md.

The high-FPS loop slept a flat 8ms after every render. display() has
already blocked on the panel's vsync by then, so that sleep was added to a
wait that had happened: ~4ms of render plus 8ms put each iteration at ~12ms
against a 10ms refresh grid, so every swap missed a refresh and the loop
settled at 50fps while asking for 125 -- with no headroom, so a further
14% of frames slipped again. It now sleeps only the remainder, with a 1ms
floor so plugin threads still get the GIL.

ScrollHelper stepped position on a wall clock at 1/scroll_delay steps per
second. Plugins set scroll_delay to the frame period, so that comparison
sat exactly on its own threshold: a frame arriving a hair early moved zero
pixels and rendered an identical frame, dirty-tracking skipped the swap, it
returned in ~2ms, and the beat repeated. No scroll_delay value tunes that
out -- a shorter delay trades stalled frames for periodic double-steps.
Both modes now accumulate elapsed time at the same configured speed, so
position stays proportional to real time.

Sub-pixel blending goes back to off by default. It renders a half-step by
mixing two adjacent columns, which on a coarse panel showing pixel-font
text alternates crisp and smeared frames and reads as shimmer -- visibly
worse than integer stepping on the hardware. Vegas mode still opts in.

disk_cache uses orjson when importable, falling back to the stdlib. Encoding
a ~1MB record drops from 14.8ms to 5.4ms end-to-end, and that work holds the
GIL while a marquee is on screen. display_manager also checksummed the whole
framebuffer twice per frame (dirty tracking, then the preview snapshot); the
snapshot now takes the checksum the caller already computed.

New src/common/scroll_config.py resolves scroll settings in one place. Five
ticker plugins each hand-rolled this and disagreed: odds-ticker ranked the
deprecated scroll_pixels_per_second above the documented scroll_speed/delay
pair, and because that key carries a schema default the documented settings
were dead for every user (ChuckBuilds/ledmatrix-plugins#408), while
ledmatrix-leaderboard read the same key only as a fallback. The resolver also
warns when a speed will not advance a whole number of pixels per refresh,
which is the property that actually determines whether a scroll looks smooth.

scripts/build_rgbmatrix_nogil.sh rebuilds the rgbmatrix binding so it
releases the GIL. Upstream declares SwapOnVSync without nogil, unlike
SetPixel/Clear/Fill beside it, so the render thread held the GIL for the
whole vsync wait and starved background threads into long uninterruptible
bursts. The script patches, builds and self-verifies into a scratch tree;
--install backs up the original and rolls back if the service does not come
back healthy.

Measured after: 100 fps locked, no stalls observed, render thread down from
51% to 19% of one core.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(display): keep the panel swap locked to vsync while scrolling

Dirty tracking skipped SwapOnVSync for byte-identical frames. That is the
right call for static content, but SwapOnVSync is also what paces the render
loop, so skipping it skips the wait for the panel: a duplicate frame returns
in ~8ms instead of ~10ms on a 100Hz panel, advances the strip only 0.8px
instead of 1.0px, and so makes the next frame more likely to repeat as well.
The effect sustains itself once it starts.

Measured over 20 minutes on a 2x128x64 chain, both scrollers configured
identically at 100 px/s:

    leaderboard   10ms x35, 11ms x3            (clean)
    odds-ticker   10ms x26, 8ms x7, 15ms x5    (~20% duplicates mid-scroll)

The duplicates were not end-of-cycle idling -- 38% of fast frames fell within
90s of a scroll completion against 35% of normal frames, a null result. The
trigger is per-frame work: odds does more of it, and more variably, so it is
first to land a frame that advances less than a whole pixel.

Pushing an identical frame costs one canvas copy. Falling out of vsync lock
costs smooth motion. Static content is untouched, because
is_currently_scrolling() expires on its own inactivity threshold -- covered
by test_stale_scrolling_state_stops_forcing_pushes so a plugin that stops
scrolling without saying so cannot pin the panel into always-push.

Also de-flakes test_snapshot_still_written_on_skip, which asserted a strict
mtime increase between two writes that can land in the same filesystem tick;
it failed about two runs in three on Windows regardless of the code under
test. The file is now backdated before the check.

156 tests pass on the Pi. Not yet confirmed by eye on the panel.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(scroll): report the frame-time tail, and stop the row-major blit

Two problems, both found by looking at the panel rather than the metric.

The frame-stats line reported ONE instantaneous frame every 5 seconds --
about 1 frame in 500 -- printed beside a 100-frame average. Both hide exactly
the fault they are used to chase: a 2ms duplicate and a 21ms double-wait
average to precisely 10ms, so a ticker stalling on half its frames still
reports a healthy "Avg FPS: 100.0". That reading cost several rounds of
chasing the wrong layer. The line now aggregates every frame since the last
log and reports median, p95, max, min, and explicit stall and skip rates
(past 1.5x the median missed a refresh; under half never reached the panel,
because dirty tracking skipped the swap so the frame never waited on vsync).

On the hardware this now reads:

    leaderboard  100.0 fps over 501 frames | median 10.00ms p95 10.05ms
                 max 10.34ms | stalls 0 (0.0%) skips 0 (0.0%)

The binding rebuild's blit patch becomes opt-in (RGB_PATCH_BLIT=1, default
off). Reordering that loop to row-major changes what a torn frame looks like:
column-major tearing shows as a vertical seam, row-major as a horizontal split
between the panel's upper and lower halves. On a 1/32 scan panel that reads as
a one-pixel fold across the middle of every panel, which is what was reported
on hardware and what went away when the blit was reverted. All of the measured
gain comes from the SwapOnVSync change, so the risky half is simply not worth
taking; the header says so.

Also fixes --install resolving its paths against $HOME, which is /root under
sudo, so it looked in /root/rgbmatrix-nogil-build and died with "no built
module found" on a machine where the build had just succeeded. It now resolves
SUDO_USER's home. Both build paths are verified on the Pi: default yields one
GIL-release site, RGB_PATCH_BLIT=1 yields two.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* feat(scroll): let users pick a crisp speed for their own panel

Whole-pixel motion was previously only available at multiples of the refresh
rate -- 100, 200, 300 px/s on a 100Hz panel. 100 px/s crosses a 256px panel in
2.6s, which is brisk for reading, and everything slower had to blend (blur) or
repeat frames unevenly (judder). There was no way to ask for 50 px/s and get
clean motion.

SwapOnVSync takes a framerate_fraction the display manager never passed. It
holds each frame for N panel refreshes; the panel keeps refreshing at its full
rate throughout, so holding costs nothing in flicker and only changes how often
a NEW image is presented. That turns 50 px/s into one whole pixel every second
refresh instead of half a pixel every refresh.

The crisp speeds are therefore refresh_hz / hold * pixels_per_frame, and that
ladder depends on the panel: a Pi Zero on a long chain has a different set of
good speeds from a Pi 4 on a short one. crisp_ladder() enumerates them and
solve_crisp() picks the best match for a requested speed.

solve_crisp weights motion quality rather than picking the numerically nearest
entry, which matters more than it sounds. Asked for 30 px/s, nearest-by-value
answers 28.6 -- 2px jumps at 14fps -- over 33.3, which is single-pixel motion
at 33fps and obviously better on the panel. The target is also clamped into the
ladder's range first, because relative error saturates near 1.0 for a target
far outside it and the quality penalty would otherwise answer "10000 px/s" with
the slowest entry.

configure() snaps to the ladder and applies the hold when given a display
manager. Without one the hold silently cannot happen and motion falls back to
fractional pixels, so it warns rather than failing quietly. set_frame_hold()
resets to 1 when scrolling stops, so one plugin's pacing cannot leak into
whatever is on screen next.

scripts/scroll_speeds.py is the user-facing part: it prints the ladder for the
configured rate, measures what the panel ACTUALLY manages (--measure, for
hardware that cannot reach its configured limit), highlights the nearest option
to a wanted speed, and demos one live. It never starts or stops the display
service itself -- doing that inside a script stranded the panel twice today.

Speeds below ~20 px/s remain stepped regardless. That is the pixel pitch, not a
software limit.

Also fixes the dirty-tracking test spy, which stubbed SwapOnVSync with a
single-argument function and would have masked the new call as a failed push,
and rewrites a configure() test that had started passing for the wrong reason:
it asserted a judder warning, which snapping now prevents, and was matching the
unrelated "hold could not be applied" warning instead.

183 tests pass on the Pi.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(scroll): tie the frame hold to the scroll, not the plugin

The hold applied in configure() never reached the panel. Plugins share one
display manager, and set_scrolling_state(False) -- fired whenever ANY other
plugin finishes its scroll -- reset the hold to 1. A hold set once at plugin
construction was therefore always gone by the time that plugin rendered.

The symptom was a log line that lied. ledmatrix-stocks reported

    Scroll configured: 50.0 px/s (1px every 2 refreshes = 50.0 fps, smooth)

while the panel measured 100.0 fps, median 10.00ms. Config, resolution and
snapping were all correct; only the pacing silently was not applied.

set_scrolling_state(is_scrolling, frame_hold=1) now carries it, so the hold
lives exactly as long as the scroll that asked for it. configure() reports the
value as ScrollSettings.frame_hold instead of applying it -- applying it behind
the caller's back could never have been right on a shared display manager.
Existing callers are unaffected; the default keeps one frame per refresh.

Verified on hardware: stocks at 50 px/s now measures

    50.0 fps over 251 frames | median 20.00ms p95 20.09ms | stalls 0 skips 0

20.00ms being exactly two refreshes, with the panel still refreshing at 100Hz
underneath so flicker is unchanged.

test_another_plugin_stopping_does_not_strand_a_hold pins the interaction that
broke this.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(scroll,cache): resolve CodeRabbit review on #523

Eight findings, all reproduced before fixing.

scroll_config.configure() read the refresh rate *after* resolve() had
already used it. resolve() fills in target_fps, pixels_per_frame and the
judder warning from that rate, so on a 60Hz panel every one of them
described 100Hz -- and with snap_to_crisp=False nothing downstream
corrected it, so set_target_fps() paced the helper to 100 FPS. The rate
is now settled first, and falls back to the global config rather than
straight to the default.

refresh_hz_from_config() used `(cfg.get("display") or {}).get(...)`,
which raises AttributeError when either level is truthy but not a
mapping -- out of a function whose whole contract is a rate or a default.

The frame-stats line reported the upper-middle sample as the median and
the 96th sorted sample as p95 of 100. Both are also thresholds (stalls
at 1.5x the median, skips at 0.5x), so the counts were biased too. The
arithmetic is now in frame_stats()/format_frame_stats(), testable
without a clock.

configure()'s docstring and docs/SCROLL_PERFORMANCE.md still said it
applies the frame hold and warns when it cannot. It deliberately does
neither since "tie the frame hold to the scroll, not the plugin"; a
caller following the old text would omit set_scrolling_state() and slow
snapped speeds would still present every refresh.

disk_cache had no policy for non-finite floats: orjson writes null,
the stdlib writes NaN/Infinity, and orjson then rejects those legacy
files so DiskCache.get deleted them as corrupt. One behaviour on both
paths now -- write null, keep legacy records readable. allow_nan=False
detects the values; the replacement walk runs only when there is one,
so the ordinary write path is byte-identical and pays nothing.

build_rgbmatrix_nogil.sh picked the build artifact with a glob piped to
`head -1`, which sorts cpython-311 ahead of cpython-313, so a stale .so
staged in from the source tree was installed as core.so while the GIL
check -- which reads the generated core.cpp, not the .so -- still passed.
It now requires the current interpreter's exact ABI name and fails
closed. Its systemctl calls were also unchecked under `set -uo pipefail`:
a failed stop left the old service running, the following start
succeeded as a no-op, and the health check reported SUCCESS for a
binding that was never loaded.

orjson floor raised to 3.11.6 for CVE-2025-67221 (unbounded recursion
in dumps); it covers the project's Python 3.10-3.13 range.

Adds test/test_cache_nonfinite_floats.py (14) plus regression tests in
test_scroll_config.py and test_scroll_helper.py. 9 of the cache tests
and 9 of the scroll_config tests fail against the pre-fix code.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_014RRtqXDCnvnY6EQwhT5CV9

* test(harness): keep the visual double's signature tied to production

Moves set_scrolling_state's frame_hold into the test double here, where
DisplayManager gains it, rather than in #534 where it arrived a PR early.
CodeRabbit flagged the #534 version correctly: a double that accepts an
argument production does not lets the call pass every harness run and
raise TypeError on the panel, which is the one failure a safety harness
exists to prevent.

The drift has now gone both ways across two branches -- double behind
production on this branch, double ahead of it on #534 -- so it is pinned
instead of remembered. test_display_double_parity.py compares the two
signatures and fails with the direction of the drift named. It reads the
files with ast rather than importing them, because display_manager
imports rgbmatrix at module scope and this check should hold on a laptop
and in CI as well as on a Pi.

Plugins begin passing frame_hold in ledmatrix-plugins#462, which is why
production and the double both need it before that lands.

Full suite: 3889 passed.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_014RRtqXDCnvnY6EQwhT5CV9

---------

Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
 2026-09-07 13:37:54 -04:00
ChuckandClaude Opus 5 a0d3e64099 fix: ten defects found validating the whole plugin fleet on hardware (#534) 
* fix(core): register tom_thumb, accept frame_hold in the test double, wire api_v3's managers

Three independent fixes found while validating every plugin on a 256x64 rig.

FontManager never registered tom_thumb even though assets/fonts/tom-thumb.bdf
ships with the core, so every plugin offering it logged "Font family
'tom_thumb' not found" (16 warnings per countdown render) and had to carry a
private loader to use a bundled font. Closes #524.

VisualTestDisplayManager.set_scrolling_state() lacked the frame_hold parameter
that DisplayManager gained, so any plugin passing it died with TypeError at
render time and failed every size. Nine plugins now make that call;
ledmatrix-stocks and ledmatrix-leaderboard were failing outright and the other
seven only passed because their scroll path was unreachable without data.
Closes #525.

api_v3 declared module-level config_manager/plugin_manager = None that nothing
ever assigned -- app.py sets the blueprint attributes, which the other 150+
call sites use. Three sites read the decoys, so /health reported the config
unreadable and the plugin system uninitialised (making "degraded" permanent and
unreachable-by-design) and /display/current fell back to a hardcoded 128x64 on
every rig. The decoys are removed rather than assigned, so a bare name is now a
NameError at test time instead of a silent None. The same function's first-call
uptime was computed from two separate clock reads and came out negative.
Closes #529.

Verified on the rig: both previously-failing plugins render, the tom_thumb
warnings are gone, /health reports "healthy" with all three checks passing, and
/display/current reports the real 256x64.

* fix(core): unique snapshot temp name, honour on-demand requests, skip empty starlark

The preview snapshot wrote through a fixed "<snapshot>.tmp". /tmp is
world-writable and sticky, and the display service runs as a different user
from the tooling, so a leftover temp owned by anyone else became unopenable
even by root -- fs.protected_regular refuses O_CREAT on a foreign file in a
sticky directory. The preview and the health check's liveness proxy then froze
until someone deleted the file by hand; on the test rig that meant 23 hours of
a healthy display reporting "hardware: stale". Now uses tempfile.mkstemp with
cleanup on failure, matching the hardware-status write a few hundred lines
above. Closes #528.

_poll_on_demand_requests read its mailbox with max_age=3600, and get() defaults
the in-memory TTL to max_age -- so the first request was pinned in memory for an
hour and every later poll returned that stale copy. No second on-demand request
was honoured until the service restarted, while the API kept returning 200.
get() already documents memory_ttl=0 for exactly this cross-process case.
The consumed request is also now deleted: leaving it on disk meant a restart
replayed the previous request, activated it, and ignored the one the caller had
just made. Closes #530.

starlark-apps returned None from display() when it has no app to show, which is
the state of every install without Pixlet and of a fresh one before any app is
added. The controller only skips on a boolean False, so that held a black panel
for the full display_duration instead of rotating on. Closes #456 (core side).

Verified on the rig: two consecutive on-demand requests with no restart between
them are both activated, where the second was previously dropped in silence.

* perf(harness): share one cache across a plugin's renders

_instantiate built a fresh MockCacheManager for every (size, mode), and that
mock is a per-instance in-memory dict, so each render was a cold start. A plugin
that fetches per game or per player re-fetched everything N times over --
baseball-scoreboard at one size took 840s for nine renders where the arithmetic
said ~72s, and at eight sizes it exceeded a 900s timeout.

The second and later renders also never exercised the cache-hit path, which is
what a running rig executes almost all of the time, so a caching regression
could not be caught here.

The cache is now built once per render_plugin_matrix call and threaded down.
The display manager stays per-render -- the bounds checking depends on that --
so only fetched data is shared.

Measured on the rig, same render counts and same goldens:
  tide-display         2s -> 1s   (32 renders)
  cricket-scoreboard  10s -> 3s   (24 renders)
No pass/fail change across tide-display, cricket-scoreboard, clock-simple,
geochron, christmas-countdown, of-the-day, web-ui-info and incoming-packages.

Closes #533.

* fix(scripts): run standalone plugin tests instead of collecting nothing

run_plugin_tests.py discovered every plugin test file and handed the lot to
pytest. Most plugin tests are standalone scripts -- module-level main() plus an
`if __name__ == "__main__"` guard, signalling through an exit code -- and pytest
collects zero items from those. The run printed how many files it had *found*,
then "no tests ran", and exited without executing any of them. On a rig with all
44 first-party plugins that is 151 of 248 files.

Files are now classified and each kind runs under the right runner: pytest for
real test modules, subprocess for scripts, honouring the 0 pass / 2 skip / 1
fail convention ledmatrix-plugins' own runner established (a script that wants a
tty or an LED matrix is a skip, not a regression).

Before:
    $ python3 scripts/run_plugin_tests.py -p countdown -d ~/LEDMatrix/plugin-repos
    Found 1 test file(s)
    collected 0 items
    no tests ran in 0.31s                      rc=0

After:
    Found 1 test file(s) -- 0 collectable, 1 standalone script(s)
    1 passed, 0 skipped, 0 failed (scripts)    rc=0

Verified across three shapes: countdown (1 script), jellyfin-now-playing and
pomodoro-timer (pytest only, 16 and 42 tests), and ledmatrix-flights (11 files
split 4 collectable / 7 scripts, all seven of which had never run).

Closes #532.

Running the flights scripts for the first time also surfaced four genuinely
failing tests there, hidden by the mirror-image bug in the plugins repo's own
runner -- filed as ChuckBuilds/ledmatrix-plugins#464 and #465.

* fix(harness): give an empty-looking mode a few frames before warning about it

check_plugin's "drew nothing but display() returned X" warning fired on a single
frame, rendered with force_clear=True, under a frozen clock. All three defeat a
scrolling plugin, whose first frame is legitimately its blank scroll-in buffer.
Across 44 first-party plugins, 60 of 76 warnings were false -- the rate at which
people stop reading a warning, which matters because the true positives are
real: a mode that draws nothing and does not return False holds a blank panel
for its whole display duration.

An apparently-empty frame is now re-driven for up to 48 more frames with
force_clear=False (force_clear means "reset the scroll", so repeating it would
redraw frame 1 for ever) and with the clock advancing -- freezegun's factory
where time is frozen, a real sleep where it is not, since scroll position is
usually a function of elapsed time. The first frame that draws content replaces
the result.

The clock is moved back afterwards. It is shared by every render in the matrix,
so time borrowed by the probe leaked into later modes and drifted their goldens
-- f1_upcoming picked up 5 spurious drifts before this was restored.

Measured on the rig:

                        empty warns          check
                        before  after
  f1-scoreboard            42      0    48 PASS / 0 FAIL, goldens intact
  ledmatrix-elections      16      0    16 PASS / 0 FAIL
  on-air                    8      8    true positive, kept
  nfl-draft                 8      8    true positive, kept
  clock-simple/geochron/    0      0    unchanged
  christmas-countdown

58 false positives gone, both true positives kept, no golden regressions. Cost
is confined to modes that really are blank: plugins that draw immediately are
unchanged (clock-simple and tide-display still 2s), while on-air -- eight
deliberately blank modes -- goes to 21s.

Closes #527.

* fix(harness): load nested schema defaults, and merge caller config at leaf level

load_config_defaults read only top-level properties. An object property carries
its defaults on its children, not on itself, so everything nested was dropped --
2,386 defaults across 37 of 44 plugins, soccer-scoreboard alone losing 539 of
565. render_plugin_matrix's comment says the plugin then "behaves like a real
install", which for most of the fleet it did not.

_defaults_from_properties now recurses. merge_config deep-merges the caller's
config onto the result so an override lands at the leaf: a shallow merge would
let -c '{"nhl": {"enabled": true}}' replace the whole nhl subtree and discard
every other nhl default, which is the same class of bug being fixed here.

Measured before/after across all 49 installed plugins on the rig: **no render
changed** -- identical PASS/FAIL counts, byte-identical output, goldens intact.
Plugins already fall back to the same values internally via config.get(key,
default), so supplying them explicitly agrees with what they were doing. The
defaults really are arriving now:

  ufc-scoreboard        9 -> 87 defaults
  ledmatrix-flights    51 -> 95
  masters-tournament   10 -> 51
  cricket-scoreboard   22 -> 50
  tide-display         12 -> 18

and hockey-scoreboard, which used to load nhl.enabled=None, now gets
nhl.enabled=True with its full display_modes block.

Caveat worth carrying: the eight plugins with the most nested config
(soccer, baseball, basketball, hockey, lacrosse, football, afl, nrl -- 1,634 of
the 2,386 dropped defaults, 68%) could not be measured. They import
src.common.sports_shared, which the test rig's core branch predates, so they
fail to load there identically before and after. Re-run this comparison against
a core that has that module before trusting the "nothing changed" result for
them; those are exactly the plugins whose renders should change most.

Closes #531.

* refactor: narrow the exception handlers this branch introduced

Codacy flagged the new code; it passes on other recent PRs, so the finding is
mine. Four of the five broad `except Exception` clauses I added were catching
far more than they needed to, which is the same shape as several bugs this
branch fixes -- hello-world's TypeError sat invisible for exactly this reason.

  freezer() / move_to() / tick()   -> (AttributeError, TypeError, ValueError)
  cache_manager.delete()           -> (OSError, AttributeError, KeyError)

The fifth stays broad and now says why: it wraps a call into a plugin's own
display(), which can raise anything, and the first frame has already rendered --
so a failure there must not turn a good result into an error.

Verified against a checkout of main: f1-scoreboard 48 PASS / 0 FAIL with 0 empty
warnings, on-air keeps its 8 true positives, clock-simple 8 PASS. geochron shows
7 golden drifts both before and after this branch, so it is not from these
changes -- its committed goldens predate #521's 1-bit text rendering.

* fix: resolve CodeRabbit review and Codacy findings on #534

CodeRabbit raised six; all six were real.

The test double had drifted ahead of production. VisualTestDisplayManager
accepted set_scrolling_state(frame_hold=...) while DisplayManager did not,
so such a call passed every harness run and would raise TypeError on the
panel -- the one failure a safety harness exists to prevent. frame_hold
belongs to the change that adds it to DisplayManager (#523), so it moves
there and the double matches main again.

The harness swallowed exceptions from re-rendered frames. _settle_loop
re-renders a mode that came back blank, to give a scroll time to draw;
returning silently on a crash meant a mode that renders one good frame
and then explodes was reported as passing. Recorded on result.error now,
keeping the captured frame so the failure stays inspectable.

starlark-apps display() returned True after _display_frame() failed, so
the controller held a dead frame for the whole display_duration instead
of rotating on. _display_frame now returns bool on all three paths.

run_plugin_tests.py used env.setdefault for PYTHONPATH and
LEDMATRIX_CORE, so an inherited value won and the subprocess imported a
different core than the one under test -- ledmatrix-plugins#467 exactly.
Prepends PROJECT_ROOT and sets LEDMATRIX_CORE unconditionally.

The on-demand mailbox is polled after every frame, ~125x/second on a
scrolling mode, and the read is deliberately uncached, so it was that
many disk reads per second to find nothing. Floored at 250ms, which is
imperceptible for a web-UI click. Consuming it also deleted whatever was
present rather than what had just been processed, so a request posted
while the previous one was in flight was thrown away and never ran; the
delete is now keyed by request_id. That narrows the window rather than
closing it -- a true atomic claim needs a primitive the cache layer does
not offer, and the code says so rather than implying otherwise.

Codacy's 2 criticals were bandit B404/B603 on the subprocess call added
to run_plugin_tests.py. Fixed interpreter, argument list, no shell;
annotated with the repo's existing nosec convention. Bandit is clean on
the file.

Adds test/test_on_demand_mailbox.py (8), test_starlark_display_contract.py
(4) and two settle cases in test_harness_empty_claimed.py. 4, 4 and 2 of
those fail against the pre-fix code. Full suite: 3961 passed.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_014RRtqXDCnvnY6EQwhT5CV9

* chore: satisfy Codacy's subprocess checks on the new test runner

Codacy runs Bandit and Opengrep (its Semgrep fork). The new
subprocess.run in scripts/run_plugin_tests.py trips three patterns, on
two different lines:

  Bandit   B404 on the import, B603 on the call
  Opengrep dangerous-subprocess-use-audit          on the run( line
           dangerous-subprocess-use-tainted-env-args on the argv line

A nosemgrep applies only to its own line, so the call line and the argv
line each need one; a single comment on the call covered neither rule
fully. Suppression is the right answer here rather than a rewrite: the
interpreter is sys.executable, the arguments are a list, and no shell is
involved, so there is nothing to word-split or expand.

Matches the pair the rest of the repo already uses for this shape --
permission_utils.py, plugin_loader.py, install_dependencies_apt.py.

Codacy: 0 new issues, up to standards.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_014RRtqXDCnvnY6EQwhT5CV9

* chore: leave visual_display_manager untouched so #523 can merge

The only change this branch made to that file was a docstring, and it
collided with #523's rewrite of the same method -- so #534 and #523 each
merged cleanly against main but conflicted with each other. Reverted to
main's text; #523 owns this method and adds frame_hold to it.

The note the docstring carried ('frame_hold arrives in #523') would have
been stale the moment #523 landed anyway. The parity test in #523 is
what actually keeps the two signatures honest.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_014RRtqXDCnvnY6EQwhT5CV9

* chore: add the Ruff suppression nosec/nosemgrep do not cover

Ruff reports S603 on the same call Bandit and Opengrep do, and none of
the three suppressions covers the others. Confirmed the precondition
first: path comes from discover_plugin_tests(), which globs test files
inside the repo, and the call is a fixed interpreter with a list argv
and no shell.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_014RRtqXDCnvnY6EQwhT5CV9

---------

Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
 2026-09-07 13:37:38 -04:00
ChuckClaude Opus 5coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com>
 		696acdbc7b feat(render_plugin): add --display-mode so multi-mode plugins can be rendered (#522) 
* feat(render_plugin): add --display-mode so multi-mode plugins can be rendered

render_plugin.py always called plugin.display(force_clear=True) with no mode.
A plugin that declares one display mode is fine, but the sports scoreboards
declare three or more and keep their per-mode state on sub-managers; their
no-argument path selects nothing and returns False, so the render came out
blank with nothing to say why. Measured on nrl-scoreboard with identical
seeded state:

  live.display() directly                 True,  1892 lit pixels
  plugin.display(display_mode="nrl_live") True,  1892 lit pixels
  plugin.display()                        False,    0 lit pixels

--display-mode passes the requested mode through. It is only passed when
asked for, so the many plugins whose display() takes no display_mode keep
working untouched, and a plugin that declares modes but does not accept the
argument degrades to its default screen with a warning rather than a
TypeError.

This is what lets the plugin READMEs show a scoreboard at all, and it also
unblocks screens like birdnet_stats and the weather plugin's hourly, daily and
almanac modes, which could previously only be described in prose.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* Only fall back when plugin display rejects display_mode

---------

Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com>
 2026-09-06 17:28:22 -04:00
ChuckandClaude Opus 5 9c0c0dc851 consolidate(web): credential exposure, secret loss, and the update path (#485) 
* fix(web): stop /config/main handing out every credential it holds

The endpoint returned the raw config to anyone who could reach the port, and
this web interface has no authentication of any kind. An unauthenticated
request against a live rig returned:

    github.api_token                40 chars
    incoming-packages.ha_token     183 chars
    jellyfin-now-playing.api_key    32 chars
    ledmatrix-weather.api_key       32 chars
    on-air.mqtt_password             8 chars
    youtube.api_key                 20 chars
    youtube-stats.api_key           39 chars

A GitHub token and a Home Assistant long-lived token among them. Anything on
that LAN could read them.

The x-secret masking the plugin config endpoints use does not reach here: this
route never consults a schema, and core keys such as github.api_token have no
schema to carry the marker. Several of the fields above *are* tagged x-secret
in their plugin's schema and were still returned in full, which is what rules
out the schema route as the fix for this endpoint.

Credential-named fields are now blanked. Matching on the name is blunt, and
for a whole-config dump that is the right default: anything named like a
credential should not leave the process, and a new plugin adding a
differently-shaped secret is covered without anyone remembering to tag it.

Blanked rather than removed, and safe to blank: POST /config/main merges into
the freshly loaded config and writes only the keys it was given, so a client
that round-trips this response cannot erase a secret it never saw. The web API
suites confirm it -- 81 passing, unchanged.

On the test that matters: the first version of this suite exercised the two
helpers and nothing else, and reverting the single line that wires the
redactor into the route passed all thirty of them. A property asserted on a
helper is not a property asserted on the endpoint, and it is the endpoint that
is exposed to the network. The added test goes through the view function, and
it does fail on that revert.

This also corrects an earlier claim of mine. I reported that GET /api/v3/config
did not expose these values; that path 404s, so the check proved nothing. The
real route is /config/main and it exposed all of them.

* fix(web): stop an unrelated config edit from erasing a plugin's secret

Saving any field on a plugin's config form destroyed that plugin's stored
credential. On a rig with a weather API key, changing the city silently
emptied the key, and the plugin stopped working at the next fetch with no
indication why.

The path had no guard at any step. The config partial masks secrets before
rendering (pages_v3.py:740), so the browser posts them back blank; _parse_value
deliberately preserves "" for optional string fields; separate_secrets routes
that "" into secrets_config, which is a truthy dict; deep_merge writes it over
the stored value; save_raw_file_content persists it.

The blank does not even need the round-trip. merge_with_defaults injects the
schema's api_key default ("") into every save, so a client that never sends
the field at all still erases it. test_secret_count_message_counts_top_level_keys
was counting exactly that injected blank as a saved secret field -- the visible
edge of the bug, pinned as expected behaviour.

remove_empty_secrets() already existed for this, with seven unit tests and a
docstring describing this precise scenario ("clients will send those empty
strings back ... so that existing stored secrets are not overwritten with
blanks"). It was never wired into a call site. This wires it into both save
paths that merge into the secrets file.

A blank now means "unchanged" rather than "delete", which is the same contract
the helper's tests already describe. The cost is that a secret can no longer be
cleared by emptying the field; clearing needs its own affordance, since a
control that erases credentials as a side effect of ordinary edits is not one.

Verified by reverting the guard: the new round-trip test then fails with the
stored key read back as ''. 262 web tests pass with it.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01STMbQE4YctTacQXfbYqKuW

* fix(web): stop dumping the config and request headers to the journal

save_main_config logged its entire POST body and the full request headers at
ERROR on every save. The body is the configuration itself, and the headers
carry the session cookie, so a routine settings change wrote both to the
journal -- at a level that guarantees they survive any sane log filter.

The lines are leftover debug output: they say "DEBUG:" in the message while
calling logging.error, and they went through the root logger rather than the
module logger, bypassing the level configured for this blueprint.

Replaced with a debug-level line recording the shape of the request, which is
the part with diagnostic value. The local `import logging` went with them; it
shadowed a module-level import that was already there.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01STMbQE4YctTacQXfbYqKuW

* fix(web): stop /config/secrets handing out every credential it holds

GET /api/v3/config/secrets returned config_secrets.json in full to anyone who
could reach the port, and this interface has no authentication. Probed against
a real rig it produced six populated credential fields: a 40-character GitHub
token, a 183-character Home Assistant token, and Jellyfin and weather API keys.
This is the second door onto the same credentials; #477 closes the first.

Masking the response alone would have been worse than the leak. The only
client fetches every secret, edits one field and posts all of them back, and
save_raw_file_content replaces the file wholesale -- so a masked GET followed
by the client's own save would write the mask over every credential the user
had not touched. That is why this was left open when the leak was found; it
needs both halves.

Read side: mask_all_secret_values(), which already existed for exactly this
endpoint -- its docstring names it -- and had never been wired to a call site.
It leaves empty values and YOUR_* placeholders alone, so a client can still
tell "set" from "not set" without being told the secret.

Write side: strip the echoed mask and blanks from the submission, then merge
onto what is stored, so "unchanged" means unchanged. The cost is that a secret
can no longer be cleared by blanking it; that wants its own affordance, since
a control that erases credentials as a side effect of saving an unrelated one
is not one.

Browser side: the token field is now left empty rather than filled from the
response. Filling it with the mask would have stored eight bullet characters
as the token the next time the user pressed Save, and filling it with the real
value is the thing being fixed. It reports whether a token is saved instead.

Verified end to end through the Flask endpoints, not the helpers. Reverting
the masking fails the leak tests; reverting the merge fails the preservation
tests; both halves are independently guarded. 278 web tests pass.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01STMbQE4YctTacQXfbYqKuW

* fix(web): stop reporting "no update" when the update check could not run

check-update returned update_available=False whenever git failed. The banner
is the only route to the update button, so a checkout git refuses to touch
looked exactly like a current one -- permanently, with nothing on screen to
act on and only a log line recording why.

The common cause is an install performed as root. scripts/install/one-shot-install.sh
clones into ${HOME}/LEDMatrix, never consults SUDO_USER, and contains no chown
at all, while its own error text suggests running the whole thing under sudo.
The result is a root-owned checkout, and on a rig this is what every git
command in it does:

    fatal: detected dubious ownership in repository at '...'

including the fetch this endpoint runs. Verified on real hardware rather than
assumed.

A failed check now reports check_failed with a message the user can act on --
for dubious ownership, the chown that fixes it. The banner shows that message
instead of hiding itself, with the update button suppressed since updating
cannot work until the cause is fixed. The success path is untouched.

This does not fix the installer, which is the real cause; it stops the symptom
being invisible. The installer needs SUDO_USER handling and a chown, and its
suggestion to run as root should go.

Reverting the endpoint change fails four of the five new tests; the fifth
guards the success path and correctly does not move.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01STMbQE4YctTacQXfbYqKuW

* fix(web): stop the installer chmod stripping exec bits on every update

git tracks five scripts as mode 644 that first_time_install.sh then chmods to
755 (start_display.sh, stop_display.sh, the two install_*_service.sh, and
one-shot-install.sh does the same to first_time_install.sh). With
core.fileMode true, the default on Linux, git reports all five as modified
from then on, in files the user never touched.

The update button stashes local changes before pulling, so it is not blocked
by this. But it never pops that stash -- stash pop and stash apply appear
nowhere in the update flow -- so the mode change is stashed away and left
there, and the files revert:

    === file modes after the update button's stash ===
      664  first_time_install.sh      <- installer had made these 755
      664  start_display.sh
      664  stop_display.sh
      664  scripts/install/install_service.sh

So every web-UI update silently strips the executable bit from the installer's
own scripts, and leaves a stash entry holding the difference. start_display.sh
and stop_display.sh stop working from the shell afterwards.

A manual `git pull --rebase` over SSH fails outright, since nothing stashes for
it: "cannot pull with rebase: You have unstaged changes". That is the likely
source of the reports, since plenty of people update that way.

Tracking the five as 755 -- what they should always have been, as the
installer chmodding them attests -- removes the spurious mode change
entirely: nothing to stash, nothing stripped, no stash entry, and manual
pulls work.

The pull also passes --autostash, for the case the code explicitly tolerates:
when the stash fails it logs a warning and pulls anyway, and that pull is what
then fails. Autostash also pops what it stashes, which the manual stash does
not.

Note that `git add -A` after `git update-index --chmod=+x` silently reverts
the index to the on-disk mode, so the modes here were set by chmodding the
files themselves.

Regression test asserts the five stay tracked executable; reverting any one
of them fails it.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01STMbQE4YctTacQXfbYqKuW

* fix(web): ask for the restart that makes an update take effect

The update button pulls new code and restarts nothing. There is no systemctl,
restart, reload or reboot anywhere in the 172-line git_pull handler -- it
stashes, pulls, installs changed requirements, re-removes plugins the user had
uninstalled, and returns "Code updated successfully."

Meanwhile both services go on running the code they loaded at boot. So the
display keeps rendering the old build, the web interface keeps serving the old
build, and the user is told the update worked. Nothing on screen suggests
otherwise, and the next reboot is what actually applies it -- whenever that is.

The affordance for this already exists: the restart-pending banner, raised
after main-config saves, with a Restart Now button wired to the display
service. A code update is a stronger reason to show it than a config save is.

The response now reports restart_required, and applyUpdate raises the banner
with wording for a code update rather than a config save. The banner's message
became a parameter and is persisted next to the flag, since it outlives the
page that raised it.

restart_required is only true when the pull actually moved HEAD. "Already up
to date" is a success too, and prompting after a no-op would train users to
dismiss the prompt unread.

This covers the display service, which is what the Restart Now button drives
and what users notice. The web interface still picks up its own new code on
its next restart; restarting it from inside a request it is serving is a
larger change than this one.

Reverting the flag fails the test that a pull which moved HEAD asks for a
restart. 290 tests pass.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01STMbQE4YctTacQXfbYqKuW

* Mask list-shaped secrets element-wise, close two vacuous tests

mask_all_secret_values treated any non-empty list as a scalar, so a
secrets file holding

  "accounts": [{"name": "a", "token": "tok-a"}, {...}]

came back as a single "••••••••". The caller could not see how many
entries existed, and the raw editor was handed a string where the file
holds an array. Recurse into lists in both _mask_value and _contains_mask.

Lists merge by replacement, not key-wise, so strip_masked_values now
drops a list outright if any element still carries the mask -- storing a
half-masked list would discard the untouched entries.

Two tests could pass without exercising what they claim to check:

- test_git_pull_resolution asserted modes only for paths git ls-files
  returned. A renamed or deleted installer target is simply absent from
  that output, so its mode was never checked. Assert every CHMODDED path
  is tracked first.
- test_config_secrets_masking never checked the POST status. A 500
  leaves the old file in place, which satisfies every assertion that
  follows. Assert 200 before reading the file back.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01STMbQE4YctTacQXfbYqKuW

---------

Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
 2026-08-21 14:30:29 -04:00
ChuckandClaude Opus 5 71739d85d1 harden: cap malloc arenas, warn on unit drift, and grant the portal's sudo commands (#476) 
* perf(systemd): cap glibc malloc arenas on the display service

Measured on a live rig 2.5 hours after start:

    RSS                          1030 MB
    Private_Dirty                 988 MB
    anonymous mappings > 10 MB       23
    largest        104, 79, 66, 63, 63 MB, on 64 MB-aligned addresses
    threads                           9
    cores                             3   -> glibc ceiling = 8 x 3 = 24 arenas

23 against a ceiling of 24, all 64 MB-aligned: these are glibc's per-thread
malloc arenas, not live objects. The data the process was actually holding
accounts for perhaps 15 MB -- the widest scroll strip observed was 35,746 x 64,
about 7 MB as RGB and the same again for its numpy mirror.

It is bloat rather than a leak: sampled four times over 135 seconds, RSS sat
between 990 and 1030 MB rather than climbing. glibc gives each allocating
thread its own arena, grows them to hold peak demand, and never gives them
back. A process that builds and drops large images across several threads is
exactly the shape that produces this.

The device had 59 MB free at the time, on 1845 MB total.

MALLOC_ARENA_MAX=2 trades a little allocator concurrency for that resident
memory. It is a tuning knob rather than a fix for a defect, so the rationale
and the measurements sit next to it in the unit file, and a test asserts they
stay there -- a bare environment variable invites removal by whoever meets it
next.

Two things this is NOT, both checked rather than assumed:

- Not an OOM problem today. A grep for "oom" in the service journal returned
  24 matches, all of which were the radar logging zoom=9 and zoom=7. The kernel
  OOM killer has not fired: dmesg has zero matches.
- Not currently capped by the unit's MemoryMax=85% either. That directive is in
  this file but absent from the unit actually installed on the rig, which
  reports MemoryMax=infinity, so nothing is enforcing a ceiling there.

The saving is unmeasured on hardware: applying it needs a service restart,
which blanks the panel, so that is the user's call rather than something to do
mid-audit. If p99 frame time regresses -- it sits at 18.4 ms against a 16.7 ms
budget for 60 FPS, so there is not much headroom -- raise the value rather than
remove it.

(cherry picked from commit 446207ffbc)

* test(systemd): pin the arena value instead of accepting a range

Review follow-up. The range check accepted 1, 3 and 4, so a change to 4 --
which hands most of the resident saving back -- passed a test whose whole
purpose is to notice that.

Pinned to the value the unit ships, in one named constant. Raising it is still
a legitimate response to a frame-time regression, but it should be a visible
edit here rather than silent drift, and the failure message says so.

Mutation-checked: changing the unit to 4 now fails.
(cherry picked from commit 73fff8d2d5)

* fix(startup): warn when an installed systemd unit has drifted from the repo's

Nothing re-applies systemd units after the first install. `git pull` -- which
is what the web UI's update button runs -- brings a new template into the
checkout, but no code in web_interface/ or src/ copies it to
/etc/systemd/system, and nothing anywhere runs `systemctl daemon-reload`. The
unit that actually runs is whatever first_time_install.sh wrote on day one.

So every hardening added to a unit is inert on existing installs, silently.
Measured on a live rig:

    installed  /etc/systemd/system/ledmatrix.service   2026-08-06
    template   systemd/ledmatrix.service               2026-08-19
    contents                                           differ

with the practical result that the MemoryMax=85% the repo's template specifies
was not being enforced at all -- `systemctl show` reported
MemoryMax=infinity. Anyone reading the template would reasonably believe the
service was capped.

Startup now compares each installed unit against its substituted template and
warns when they differ, naming install_service.sh as the remedy.

A warning, not an error, and deliberately not a silent rewrite: editing files
under /etc and restarting services is the installer's job, not something a
display process should do to a machine while it is booting. Making it fatal
would also brick every development checkout whose unit is legitimately absent
or hand-edited.

Comparison ignores comments, blank lines and ordering. The template carries
explanatory comments the installed copy will not have, and systemd does not
care about order within a section, so a literal comparison would warn on every
boot and be ignored within a week.

Mutation-checked three ways: never reporting drift fails, making it fatal
fails, and -- after the first attempt missed it -- comparing raw text now fails
too. That last gap is worth noting: the comment-insensitivity tests originally
exercised the helper directly, so a comparison that stopped calling the helper
passed them all. The test that catches it goes through _validate_systemd_units.

29 startup-validator tests pass.

(cherry picked from commit cf521bdfd8)

* fix(install): grant the sudo commands the captive portal actually runs

The installers write two allow-lists, /etc/sudoers.d/ledmatrix_web and
ledmatrix_wifi. Anything the code runs under sudo that is not in one of them
needs a password, which a service cannot supply, so the call fails.

Five commands were being run and none of them granted:

    sysctl -w net.ipv4.ip_forward=0|1     wifi_manager.py:788, 883
    nft add|delete table ip ledmatrix     wifi_manager.py:835, 895
    rfkill unblock wifi                   wifi_manager.py:1811
    iptables ...                          wifi_manager.py:796, 813, 818, 871
    mkdir -p .../dnsmasq-shared.d         wifi_manager.py:922

Together these are the captive portal: unblock the radio, bring up the AP,
add the redirect, turn on forwarding, and undo all of it afterwards. Without
the grants a hardened install would associate clients to the access point and
then fail to route them.

Why it has gone unnoticed: a stock Raspberry Pi image ships
/etc/sudoers.d/010_pi-nopasswd granting the default user

    <user> ALL=(ALL) NOPASSWD: ALL

which satisfies every one of these regardless of what the allow-lists say.
Confirmed on a live rig -- `sudo -n -l` permits sysctl there, and the blanket
rule is why. The allow-lists are effectively decorative on a default image and
only start mattering once that rule is removed or the service runs as another
user.

test_sudo_allowlist_covers_calls.py extracts every argv-style sudo call in
src/ and web_interface/ and asserts an installer grants it, so the next command
added without a rule fails here rather than on someone's hardened box.

Getting that test honest took three passes, each worth recording:

- Matching the literal "systemctl" against rules written as
  `$SYSTEMCTL_PATH enable ...` reported six gaps that did not exist. Binary
  path variables are now normalised before comparing.
- Scanning the whole installer let `NFT_PATH=$(command -v nft)` -- a variable
  definition, not a grant -- satisfy the check on its own, so deleting the
  actual nft rules still passed. Only NOPASSWD lines are considered now.
- `sudo -n <tool>` reported "-n" as the binary. sudo's own flags are skipped.

Each of the five grants is individually mutation-checked: removing any one
fails the suite.

(cherry picked from commit a372b43cd1)

* fix(install): drop the iptables wildcard, and pin each grant properly

Review follow-up. Two findings, both right, and the first is a hole I opened
myself.

`NOPASSWD: iptables *` is a root shell for the web user by another name.
`iptables --modprobe=/path/to/anything` runs that path as root, so a wildcard
grant on iptables escalates rather than restricts. I added that rule while
fixing a permissions gap, which is a worse outcome than the gap. It is gone,
and a test now fails on any trailing-wildcard grant to a tool that can execute
another program -- iptables, nft, tcpdump, find, awk, sed, perl, python, env.

The other finding: checking only the binary made the coverage test far weaker
than it looked. With `sysctl` present anywhere in the allow-list, deleting the
`net.ipv4.ip_forward=0` grant still passed -- and the portal would then be
unable to restore forwarding on teardown. Each required command is now matched
in full, and each is mutation-checked individually, including that exact
single-line case.

Scope pulled in deliberately. The first version of this test tried to assert
that *every* sudo call in the codebase is granted. Run honestly, it showed the
portal also runs iptables, nft, `ip addr`, `ip link` and `cp` with arguments
built at runtime -- an interface name, a port. Those cannot be granted safely
in a sudoers file: the rule needs a trailing wildcard, and that is the
escalation above. Closing that half needs a privileged helper that builds the
rules itself and takes only an interface and a port, granted the way
safe_plugin_rm.sh already is. That is a design decision, not a one-line grant,
so the test now pins the four commands this change actually grants and the
docstring says plainly what it does not cover.

Better a narrow test that is true than a broad one that is not.

(cherry picked from commit 500cfbc9f4)

* fix(install): pin PATH, keep unit order, and tighten the sudoers assertions

Three review findings, all correct.

The installer resolved binaries through an inherited PATH and wrote whatever
it found into sudoers as NOPASSWD grants. first_time_install.sh re-execs
itself with `sudo -E`, which preserves the caller's environment, so a writable
directory early in PATH turned a compromise of the low-privilege web user into
permanent root -- via a file the installer itself wrote. PATH is now pinned to
the system directories before anything is resolved, and every resolved binary
must be root-owned and unwritable by anyone else before it reaches the
sudoers file.

_unit_body() sorted a unit's lines before comparing. Order is not noise in a
systemd unit: repeated ExecStartPre=/ExecStartPost= run in the order they
appear, and a directive that moves between [Unit], [Service] and [Install]
means something different where it lands. The drift check reported no drift
for units that had genuinely changed. Order is preserved now.

Two of that check's own tests asserted the wrong thing --
test_reordered_directives_are_not_drift said so in its name -- and are
inverted, with a second covering a directive moved between sections. The
cosmetic-difference test now varies comments, blank lines and indentation,
which is what the installer actually drops, rather than reversing the file.

The sudoers assertions matched command prefixes, so
`sysctl -w net.ipv4.ip_forward=0 *` satisfied the requirement while granting
the caller arbitrary trailing arguments as root. They are exact now. The
wildcard check also normalises ${NFT_PATH} the same way as $NFT_PATH; the
brace is not a word boundary, so that spelling was skipped entirely.

Verified by reintroducing each: a widened required grant fails the exact
match, `${NFT_PATH} *` fails the wildcard check, and require_trusted_binary
refuses a non-root-owned, world-writable, or missing binary.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01STMbQE4YctTacQXfbYqKuW

---------

Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
 2026-08-21 14:29:45 -04:00
ChuckandClaude Opus 5 cc258aaffd fix(install): tag the secondary installer's journalctl grants too (#491) 
Review was right on all three counts, and the first is the one that matters:
scripts/install/configure_web_sudo.sh writes the same three wildcard
journalctl rules as first_time_install.sh and none of them carried NOEXEC. So
this PR closed the pager escape on one installer path and left it open on the
other, which is close to no fix at all -- a rig configured through that script
still hands out a root shell via less's "!command".

The test could not have caught it, for two independent reasons. INSTALLERS
did not list the file. And even listed, _grant_lines() kept the raw source
line: that installer echoes its rules, so each one ends in a quote rather
than the wildcard, and the trailing-* check skipped every one of them. Either
alone would have hidden it.

Both fixed: the file is covered, and an echoed rule is unwrapped to the
sudoers line it actually emits.

The selector test now covers -t ledmatrix as well. It asserted only the two
-u forms, so deleting the -t rule would have passed.

Verified by removing NOEXEC again from the secondary installer: four of the
six tests fail, where before the suite passed with the vulnerability present.


Claude-Session: https://claude.ai/code/session_01STMbQE4YctTacQXfbYqKuW

Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
 2026-08-21 14:28:23 -04:00
ChuckandClaude Opus 5 9083df9f5c fix(pixlet): resolve the release tag correctly when downloading (#461) 
* fix(pixlet): resolve the release tag correctly when downloading

Starlark apps render through the pixlet binary, and the installer that
fetches it silently produced nothing, so every app failed with "Pixlet
not available - Starlark apps will not work".

Two compounding defects:

The version lookup parsed the wrong token. GitHub returns the release
JSON on a single line, so `grep '"tag_name"'` matches the whole document
and the greedy `sed 's/.*"([^"]+)".*/\1/'` captures the LAST quoted
string in it. That resolved to "mentions_count", giving a download URL
for a release that does not exist. The `[ -z "$PIXLET_VERSION" ]`
fallback never fired, because the value was not empty -- just wrong.

And `curl -L -o` without `-f` writes a 404 body to the file and exits 0,
so the download was reported as successful and the first sign of trouble
was tar complaining "not in gzip format" about a page of HTML:

    → Downloading linux-arm64...
      Extracting...
    gzip: stdin: not in gzip format
    ✗ Failed to extract archive: .../pixlet_mentions_count_linux-arm64.tar.gz
    Download complete: 0/1 succeeded

Now the tag field is matched directly and the value taken from it, and
the result is checked for a version shape rather than merely being
non-empty -- a wrong-but-non-empty value is exactly what made this
silent. curl gets -f so an HTTP error is a failure, and the archive is
gzip-tested before extraction, since a proxy can return 200 with an
error page.

Verified on an arm64 rig: v0.53.1 resolved, 1/1 downloaded, the binary
runs, and the plugin's own detection finds it at
bin/pixlet/pixlet-linux-arm64.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Udr6MfaFLUPhX5Fgo67Jf5

* fix(pixlet): anchor the version check, and don't echo response bytes raw

Both CodeRabbit findings were valid.

The shape check accepted partial matches, so "v0.53garbage", "0.53" and
"v0.5" passed it and built a download URL for a release that cannot exist
-- the failure the check was added to stop, just one step later. Anchored
at both ends now. Every tronbyt/pixlet release to date is vX.Y.Z (all 38
verified against the API), with an optional suffix left for a future -rc.1
or +build tag.

The invalid-response diagnostic printed bytes straight from whatever
answered the request. NUL and newline were filtered but escape, carriage
return and backspace were not, so an error page could rewrite the output
or bury it in a CI log. Non-printable bytes are stripped and it goes
through printf. CodeRabbit suggested hex-encoding the lot; printable
characters are kept instead, because "<!DOCTYPE html>" is the diagnostic
-- hex would make the line safe and useless.

Also corrected the comment above the parse. It asserted GitHub returns
this JSON on a single line; the API is pretty-printed by default, and I
could not get a single-line response from two machines across five header
variants. The single-line case is real (it is what produces
"mentions_count", and the failing device's error named
pixlet_mentions_count_linux-arm64.tar.gz), but it is a shape to be robust
against, not a constant. As written the comment invites the next reader to
check by hand, see pretty JSON, and conclude the fix was unnecessary.

Tests drive the real script with a stubbed curl: the tag resolves from
both response shapes, non-release values fall back, an HTTP error is
reported as a download failure rather than surfacing later as a tar error,
a non-archive body is rejected before extraction, and the diagnostic
cannot carry control bytes. The stub honours -f the way real curl does --
without that, the HTTP-error test passed against the old script too, since
both end at 0/1 and only the reporting layer differs.

Mutation-checked: 10 of the 16 fail against the pre-fix script, and the 5
covering these two findings fail against this branch's previous state.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01STMbQE4YctTacQXfbYqKuW

---------

Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
 2026-08-17 15:35:02 -04:00
ChuckandClaude Opus 5 f887063434 test(harness): flag a mode that draws nothing without reporting it (#447) 
The controller skips a mode whose display() returns False and treats
anything else -- including None -- as "content was shown". A mode that
draws nothing and does not return False is therefore never skipped, and
because a mode switch clears the panel first, it sits on a blank screen
for its whole display duration. Two sports plugins shipped exactly that.

The harness rendered those modes and passed them, because it called
display() and discarded the result. Capture it, and warn when a render
produced no lit pixels while claiming content.

Warn-only by default, and deliberately so: a scroll mode's first frame
is legitimately its blank scroll-in buffer, which is 42 of these on the
F1 scoreboard alone. Plugins whose modes are known to draw on their
fixture data can opt into failing via harness.json {"empty_check":
"strict"}, matching how the fill check is staged.

Worth being clear about the limit: this only sees what the fixtures
render. It would not have caught the sports bug, whose fixture seeds
games so the empty path never renders -- that needs the source-level
gate in the plugins repo. What it does catch is the same mistake in any
plugin whose empty state the harness does happen to reach, which is
coverage there was none of before.


Claude-Session: https://claude.ai/code/session_01Udr6MfaFLUPhX5Fgo67Jf5

Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
 2026-08-10 08:17:57 -04:00
ChuckandClaude Fable 5 d9683e28be Codebase audit: fix shipping bugs, remove verified-dead code, repair doc drift, add regression guards (#438) 
* fix(web): implement delete_cached so the font catalog cache actually invalidates

api_v3.py's font upload/delete handlers import delete_cached from
web_interface.cache, but the function was never defined. The surrounding
except ImportError silently swallowed the failure, so the fonts_catalog
cache entry survived uploads/deletes and newly uploaded fonts did not
appear until the TTL expired or the service restarted.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SXb4mKcAkVaxkeTb3YnAdr

* fix(web): remove dead weather/stocks partial routes that returned 500

The partial dispatcher still routed 'weather' and 'stocks' to loaders
rendering v3/partials/weather.html and stocks.html — templates that no
longer exist since weather and stocks became store plugins. Requesting
either partial raised TemplateNotFound, which the catch-all turned into
a 500. No template or JS references these partials (the only 'weather'
hit in the front end is a plugin-store category filter option), so the
branches and both loader functions are removed; unknown partials now
fall through to the existing 404 handler.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SXb4mKcAkVaxkeTb3YnAdr

* fix(deps): align contradictory psutil/Flask-Limiter/freetype-py pins

requirements.txt's optional-install comment recommended psutil>=5.9,<6.0
while web_interface/requirements.txt hard-requires >=6.0,<7.0 — anyone
following the comment ends up with an unsatisfiable pair. The comment now
recommends the same range the web interface requires (all psutil APIs
used — Process, boot_time, cpu_percent, disk_usage, virtual_memory — are
stable in 6.x). Flask-Limiter gains the same <4.0 cap in both files and
freetype-py the same >=2.5.1 floor.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SXb4mKcAkVaxkeTb3YnAdr

* fix(config): add template keys the code already reads

display.hardware gains pixel_mapper_config, row_address_type,
multiplexing and panel_type (read at display_manager.py with these exact
fallbacks — users on non-standard panels previously had no way to
discover them from the template). vegas_scroll gains
frame_based_scrolling and scroll_delay, the only two of its 27 keys the
template omitted (read in src/vegas_mode/config.py). plugin_system gains
development_mode, which the web UI reads and writes but the template
never declared.

Every added value is byte-identical to the code-side .get() fallback, so
ConfigManager._migrate_config() merging these keys into existing user
configs cannot change behavior on any installed device.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SXb4mKcAkVaxkeTb3YnAdr

* fix(scripts): repair broken sys.path setup in utility scripts

clear_cache.py and download_nba_logos.py pointed sys.path at a 'src'
directory relative to the script's own folder (scripts/utils/src and
scripts/src — neither exists), so both crashed on import; they now insert
the project root and import via the src package like the other scripts.
debug_web_manual.py resolved 'project root' to scripts/debug/ instead of
two levels up. fix_nhl_cache.sh is removed: it used Python docstring
syntax in a bash script and invoked clear_nhl_cache.py, which does not
exist anywhere in the repo — it cannot ever have worked in its current
location.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SXb4mKcAkVaxkeTb3YnAdr

* docs: correct stale file:line references and the loader-fallback contradiction

CLAUDE.md and .cursorrules disagreed about plugin-directory fallback
behavior; the code (SchemaManager.get_schema_path) probes plugins/
BEFORE plugin-repos/, and the main discovery path has no fallback at
all — both files now describe the real behavior, preferring symbol names
over line numbers so the references rot slower. REST_API_REFERENCE.md
pointed at app.py:144/:607 for mounts that live at :199/:799 and counted
92 routes where there are 94. PLUGIN_ARCHITECTURE_SPEC.md's historical
banner gains a note that its example imports
(src/plugin_system/base_classes/*_plugin.py) never shipped — the real
base classes are src.base_classes.sports.SportsCore and
src.base_classes.hockey.Hockey.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SXb4mKcAkVaxkeTb3YnAdr

* docs: fix broken links, phantom script references, and stale CI description

Repairs every broken relative link in active docs (targets renamed or
archived long ago: PLUGIN_DEVELOPMENT.md -> PLUGIN_DEVELOPMENT_GUIDE.md,
API_REFERENCE.md -> REST_API_REFERENCE.md, PLUGIN_STORE_USER_GUIDE.md ->
PLUGIN_STORE_GUIDE.md, plugin_docs/ dir, TROUBLESHOOTING_QUICK_START.md,
and MIGRATION_GUIDE's README link that silently resolved to the docs
index instead of the project README). Replaces commands invoking scripts
that do not exist (scripts/update_stats.py, validate_registry.py,
check_updates.py, fix_permissions.sh) with the real tooling, and
rewrites HOW_TO_RUN_TESTS.md's CI section, which described a
security-audit workflow that was never committed and a pytest workflow
'queued to land' that landed long ago as test.yml.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SXb4mKcAkVaxkeTb3YnAdr

* docs: complete the docs index and refresh the web interface file tree

docs/README.md's own policy says every page must be linked from the
index, yet five weren't — including the entire skin system
(SKIN_SYSTEM.md, CREATING_SKINS.md), ADAPTIVE_LAYOUT.md,
plugin-safety-harness.md and SPORTS_UNIFICATION.md. Each is now listed
in the section it belongs to, and PLUGIN_ARCHITECTURE_SPEC.md is marked
historical in the index (the doc itself already carries the banner).
web_interface/README.md's static/v3 tree showed only app.css/app.js;
it now reflects the actual contents.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SXb4mKcAkVaxkeTb3YnAdr

* chore: remove dead modules confirmed unused in-repo and across all store plugins

- src/common/cli.py: imports a 'ledmatrix_common' package that exists
  nowhere (not in this repo, any requirements file, or the plugin
  monorepo), so it cannot ever have run; its README section claimed
  scripts/dev/* used it, which was also untrue.
- src/web_interface/logging_config.py: zero callers — the web app uses
  web_interface/logging_config.py (a different module), and nothing
  imports the src copy.
- handle_errors decorator in src/web_interface/error_handler.py: zero
  call sites (the module's response helpers stay — they are used).
- ConfigManager.get_clock_config(): reads a 'clock' config key that no
  longer exists anywhere; only caller was its own unit test.

Deliberately kept despite zero in-repo callers: DisplayError,
src/common/config_helper.py and display_helper.py — all documented as
plugin-facing API (docs/PLUGIN_ERROR_HANDLING.md, src/common/README.md),
and third-party plugins outside the official monorepo cannot be
enumerated. Verified against a fresh clone of ledmatrix-plugins (43
plugins): zero references to any removed symbol.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SXb4mKcAkVaxkeTb3YnAdr

* chore: remove manager-era NBA test files and one-off debug scripts

The four test_nba_*.py files imported nba_managers, leaderboard_manager
and odds_manager — top-level modules deleted when sports displays became
plugins — inside try/except blocks that swallowed the ImportError, so
they passed while exercising nothing. test_nba_data_structure.py and
debug_nba_api.py (a diagnostic script living in test/) made live ESPN
API calls rather than testing repo code. None were enrolled in CI.

scripts/debug/direct_fix_imports.py and check_imports.py were one-shot
artifacts that edited/inspected a hardcoded ~/LEDMatrix/web_interface/
app.py to fix an import problem solved long ago; nothing references
them.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SXb4mKcAkVaxkeTb3YnAdr

* chore: remove generate_report.py, which aggregates artifacts of CI jobs that do not exist

The script's only function is to merge JSON artifacts
(bandit/semgrep/pip-audit/safety/gitleaks results) produced by a
security-audit workflow that was never committed —
.github/workflows/ has no such jobs, so there is nothing for it to
aggregate and no way to run it usefully. Its siblings stay:
prove_security.py and audit_plugins.py both run standalone (verified),
and .codacy.yml stays because the Codacy service (README badge) reads it
server-side without a workflow file.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SXb4mKcAkVaxkeTb3YnAdr

* fix(web): load the three widget scripts store plugins already declare

time-picker.js, file-upload-single.js and plugin-file-manager.js
register widgets that installed store plugins reference in their config
schemas (countdown uses x-widget: time-picker and file-upload-single;
of-the-day uses plugin-file-manager), but base.html never included the
scripts. plugin_config.html renders such fields as an empty container
that polls LEDMatrixWidgets.get(...) on a 50ms loop forever, so those
plugin config fields appeared permanently blank. The audit initially
flagged these files as dead code; the monorepo cross-check proved the
opposite — they were unreachable, not unused.

example-color-picker.js (the documented custom-widget example) gains an
explicit warning that including it in base.html would shadow the
built-in color-picker widget.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SXb4mKcAkVaxkeTb3YnAdr

* chore: drop the legacy youtube block from the secrets template

No code reads a top-level youtube secrets key: the youtube-stats plugin
receives its API key namespaced under its own plugin id (declared via
x-secret in its config schema), like every other store plugin. The key
survives only in state_reconciliation.py's non-plugin-key exclusion set,
which stays — existing installs still carry the key in their generated
config_secrets.json, and the exclusion prevents it from being
misclassified as a plugin config. New installs simply stop being asked
for a YouTube API key they have nowhere to use.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SXb4mKcAkVaxkeTb3YnAdr

* chore(deps): remove packages nothing imports, declare direct imports, move mypy to test deps

Removed from requirements.txt: python-socketio, python-engineio,
websockets, websocket-client — zero imports anywhere in this repo, and
the one store plugin that needs Socket.IO (ledmatrix-music) declares it
in its own requirements.txt, which the plugin store installs. Removed
the same quartet plus timezonefinder, geopy, google-auth-oauthlib,
google-auth-httplib2, google-api-python-client, unidecode, icalevents,
python-dateutil, flask-wtf and the werkzeug pin from
web_interface/requirements.txt — all leftovers from the deleted built-in
weather/calendar/music displays (flask-wtf was doubly dead: app.py
explicitly disables CSRF and sets csrf=None). scripts/
install_dependencies_apt.py, which mirrors these lists for the
first-time installer, drops the same packages.

Added: urllib3 (imported directly in four core modules), jinja2 and
markupsafe (imported directly in pages_v3.py) — previously reachable
only as transitives. mypy moves from runtime requirements to
requirements-test.txt.

Verified in a fresh venv: all four requirements files co-install, pip
check is clean, the full CI-enrolled suite (907 tests) and a Flask boot
smoke pass with the trimmed dependency set.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SXb4mKcAkVaxkeTb3YnAdr

* refactor: single canonical DateTimeEncoder

src/cache_manager.py and src/cache/disk_cache.py each defined an
identical DateTimeEncoder (datetime -> ISO-8601). The disk_cache copy is
the only one actually used for serialization; cache_manager now
re-exports it instead of defining a twin, so the two can never silently
diverge. Import compatibility is preserved — from src.cache_manager
import DateTimeEncoder still works and is the same class object.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SXb4mKcAkVaxkeTb3YnAdr

* docs(code): document deliberate duplicates instead of merging them

The audit surfaced several near-duplicate implementations that turned
out to be either deliberate forks or behaviorally different — merging
any of them would risk changing behavior on installed devices, so each
now carries an explicit comment stating the relationship:

- VisualDisplayManager: headless fork of DisplayManager; header now
  lists the ~15 mirrored methods and warns that DisplayManager changes
  must be mirrored.
- normalize_abbreviation: LogoDownloader's version (called directly by
  nine scoreboard plugins) replaces filesystem-unsafe characters;
  LogoHelper's strips spaces. Logo filenames on existing installs
  depend on both behaviors staying put.
- The two PluginTestBase classes: the shipped one is plugin-author
  API, the repo's own richer harness lives in test/plugins/ — now
  cross-referenced.

Also verified (no change needed): ConfigManager's backup/rollback
methods genuinely delegate to AtomicConfigManager, and SportsCore
already delegates _read_bdf_native_size to FontManager.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SXb4mKcAkVaxkeTb3YnAdr

* docs: add a unified configuration reference

There was no single place documenting what lives in config.json —
display.* keys were scattered across README sections, vegas_scroll lived
in ADVANCED_FEATURES.md, and dim_schedule, display.double_sided,
sync.follower_position, plugin_system.development_mode and the four
newly-templated hardware keys were documented nowhere. CONFIG_REFERENCE.md
now lists every template key plus the code-read-only keys, each with
type, default, and the code location that reads it, and explains the
secrets file's plugin-id namespacing. Linked from the docs index.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SXb4mKcAkVaxkeTb3YnAdr

* docs: bring the README's feature tour into the plugin era

The Core Features section still presented clock/weather/sports/stocks/
music displays as built into the project, when all of them are store
plugins installed from the ledmatrix-plugins monorepo — only
starlark-apps and web-ui-info ship in this repo. The intro now says so
(the showcase itself is unchanged; those are real displays available in
the store). The display_durations reference drops its built-in-calendar
example in favor of plugin-id keys, and the Configuration section links
the new CONFIG_REFERENCE.md.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SXb4mKcAkVaxkeTb3YnAdr

* docs: archive the custom-icons status report, cross-link config docs, document assets/

PLUGIN_CUSTOM_ICONS_FEATURE.md was a 'What Was Implemented' status
report duplicating the actual guide (PLUGIN_CUSTOM_ICONS.md) — moved to
docs/archive/ per the docs index's own policy. The overlapping
plugin-config docs keep their content but PLUGIN_CONFIG_ARCHITECTURE.md
now states up front which doc is canonical for which purpose.

assets/README.md is new and load-bearing: assets/stocks, weather,
news_logos and broadcast_logos have zero references in this repo's code,
which makes them look deletable — but store plugins (ledmatrix-stocks,
ledmatrix-weather, news, odds-ticker) resolve those exact paths at
runtime against the install directory. The README records that evidence
so a future cleanup doesn't break installed plugins.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SXb4mKcAkVaxkeTb3YnAdr

* test: add regression guards for the bug classes fixed in this PR

Three lightweight static checks, all enrolled in CI's unit-test
allowlist along with the new web-cache test:

- test_template_targets.py: every literal render_template() target must
  exist (would have caught the weather/stocks partial 500s at commit
  time).
- test_widget_scripts.py: every widget JS file must be script-included
  in base.html or explicitly allowlisted with a reason (would have
  caught the unloaded time-picker/file-upload-single/plugin-file-manager
  widgets), and allowlisted files must NOT be included (prevents the
  example widget from shadowing the real color-picker).
- test_doc_links.py: relative markdown links in active docs must
  resolve (docs/archive/ exempt).

Each guard was verified to fail against the pre-PR tree and pass now.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SXb4mKcAkVaxkeTb3YnAdr

* fix(deps): restore the werkzeug version floor

Commit 1ec22db removed the werkzeug>=3.1.6,<4.0.0 pin along with the
genuinely-unused packages, but this one was a version floor on Flask's
transitive dependency, not a phantom: Flask 3.1.3 itself only requires
werkzeug>=3.1.0, so dropping the pin let fresh installs resolve
3.1.0-3.1.5. Restored with a comment explaining why it exists.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SXb4mKcAkVaxkeTb3YnAdr

* fix(web): route pixel_mapper_config into display.hardware; guard time-picker registration

pixel_mapper_config was the only display.hardware key absent from both
the display_fields detection allowlist and the hardware write loop in
the settings save path. No form posts it today, but if one ever did the
key would fall through to the generic handler and land at the TOP level
of config.json — where state_reconciliation would mistake it for a
missing plugin id and loop auto-repair attempts (the failure class the
'github'/'youtube' exclusion comment documents). It now round-trips
into display.hardware like its siblings.

time-picker.js gains the same LEDMatrixWidgets-undefined guard its two
sibling widgets already have; correct today only via defer ordering.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SXb4mKcAkVaxkeTb3YnAdr

* chore: align installer leftovers with the dependency cleanup

first_time_install.sh's fallback secrets heredoc (used only when the
template is missing) still wrote the legacy youtube block — now matches
the template (github only). install_dependencies_apt.py drops the
IMPORT_NAME_MAP entries for packages no longer in its install lists and
a stale google-api reference in a docstring.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SXb4mKcAkVaxkeTb3YnAdr

* fix: address CodeRabbit review findings

Verified each finding against the code; fixes for the valid ones:

- install_dependencies_apt.py: the installer listed 'freetype', but the
  declared dependency is freetype-py — an apt miss would pip-install the
  wrong PyPI package. Now installs freetype-py with an import-name
  mapping (pre-existing bug, surfaced by the review).
- api_v3.py: pixel_mapper_config is validated as a string before being
  saved to display.hardware (JSON callers could previously store an
  object/list the matrix library can't use).
- .cursorrules: the Plugin Loading Process and File Organization
  sections still said discovery scans plugins/ — now consistent with the
  corrected overview (configured directory, default plugin-repos/).
- README.md: removed the stale '(except the core calendar)' claim — no
  core calendar exists in src/ — and qualified the plugin inventory
  (official plugins in the monorepo; third-party from their own repos).
- CONFIG_REFERENCE.md: hardware_mapping now shows the code fallback
  (adafruit-hat-pwm) alongside the template value.
- PLUGIN_REGISTRY_SETUP_GUIDE.md: check_plugin.py takes --plugin, not a
  positional id.
- scripts/fix_perms/fix_*.sh: exec bits set so the documented
  'sudo ./...' invocations work.
- Guard tests hardened: template guard now catches multi-line
  render_template() calls; widget guard parses actual <script> src
  values and fails if the widgets dir goes missing; type hints and
  docstrings added per repo coding guidelines.

Skipped with reasons (noted on the PR): limit_refresh_rate_hz 100-vs-90
is documented as intentional in CONFIG_REFERENCE.md; the psutil comment
already names the enforcing manifest; docs/archive/ findings are out of
scope per the docs policy (archive may rot).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SXb4mKcAkVaxkeTb3YnAdr

* chore: remove Cursor IDE tooling, consolidate its guidance into CLAUDE.md

The maintainer no longer uses Cursor. .cursorrules, .cursorignore and the
.cursor/ tree (rules, plugin templates, a parallel 751-line plugins
guide) are removed; measurement showed near-zero literal overlap risk —
the canonical content already lives in docs/. Unique guidance worth
keeping moved before deletion:

- CLAUDE.md gains the dev workflow (dev_plugin_setup.sh, dev_server.py,
  run.py -e, check_plugin.py), the plugin-secrets namespacing contract,
  and the no-draw_image()/paste-onto-PIL pitfall.
- PLUGIN_DEVELOPMENT_GUIDE.md absorbs the plugin version-management
  rules (pre-push hook install, SKIP_TAG, version resolution order) that
  its own text previously linked out to .cursorrules for.
- The one completed plan doc (.cursor/plans/) is archived to
  docs/archive/ per the docs policy rather than deleted.

One of the deleted rule files (sports-managers.mdc) targeted
src/*_managers.py globs that have matched nothing since the plugin
migration.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SXb4mKcAkVaxkeTb3YnAdr

* chore: second-pass cleanup — dead installer branch, broken script, orphaned JS, misfiled test deps

- first_time_install.sh: removed the pip fallback branch that installed
  from requirements_web_v2.txt — a file that has not existed since the
  v2 web interface was removed (the branch always printed its own
  'not found; skipping' warning).
- scripts/remove_plugin_backups.sh deleted: its PROJECT_ROOT resolved to
  the repo's PARENT directory, and its verify_submodules() checks for
  plugin submodules from an era before plugins moved to the store — it
  could never have worked from its current location.
- plugins_manager.js: removed three functions with zero call sites
  anywhere (addKeyValuePair, formatCommit, togglePasswordVisibility) —
  verified against all templates, all JS, and the dynamic window[name]
  dispatch sites, which resolve widget-registry keys only. Also replaced
  base.html's misleading 'Legacy ... during migration' label: the file
  is deliberately loaded last and provides the LIVE implementations of
  seven window.* plugin actions that shadow same-named definitions in
  app.js/app-shell.js.
- pytest/pytest-cov/pytest-mock moved from runtime requirements.txt to
  requirements-test.txt (CI already installs both files; the installer's
  line-by-line loop simply installs three fewer packages on devices; no
  store plugin declares pytest). HOW_TO_RUN_TESTS.md updated.
- scripts/add_defaults_to_schemas.py and analyze_plugin_schemas.py
  scanned the empty legacy plugins/ dir — now scan plugin-repos/.

Verified: fresh venv installs all four requirements files with pip check
clean and pytest available; bash -n on the installer; node --check on
the JS; widget/cache guard tests green.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SXb4mKcAkVaxkeTb3YnAdr

* docs: correct semantically stale content across the user and developer guides

A second-pass content audit checked the guides' substantive claims
against the code (the first pass only fixed mechanical drift). Fixes:

- GETTING_STARTED: described booting a prebuilt SD image and seeing
  default clock/weather plugins — neither exists. Now documents the real
  install (Pi OS Lite + one-shot installer / first_time_install.sh) and
  that displays come from the Plugin Store. Duration and ordering
  instructions moved to the Rotation tab where the controls actually
  live.
- WEB_INTERFACE_GUIDE: three whole tabs were undocumented (Rotation,
  Backup & Restore, Tools) and the Display tab's Vegas Scroll section
  was unmentioned. Fonts overrides are per display element (not per
  plugin); Logs has an Auto-scroll checkbox (not a Pause button); the
  aspirational keyboard-shortcut list and no-JS claim removed.
- TROUBLESHOOTING: the hand-written service-file template (wrong user,
  wrong ExecStart, dropped the autostart gate) replaced with the real
  systemd/ units + install scripts; recovery steps no longer copy
  placeholder units verbatim; WiFi curl endpoint corrected to /api/v3/;
  cache-clearing advice now targets the real cache locations.
- ADVANCED_FEATURES: removed a false claim that CacheManager has no
  delete(); fixed two example snippets that raise TypeError
  (BackgroundDataService and get_config_file_mode signatures); fixed
  cache paths, a 5-minute TTL that is actually 1 hour, and the vegas
  table now links the complete 26-key reference.
- EMULATOR_SETUP_GUIDE: documented run.py flags that don't exist
  (--plugin/--test-plugins) removed in favor of dev_server.py and
  check_plugin.py; shipped emulator config values corrected (browser
  adapter default on :8888, not pygame).
- PLUGIN_QUICK_REFERENCE: drag-and-drop reordering is shipped, not
  'not yet supported'; discovery-fallback and registry-repo claims
  corrected. PLUGIN_API_REFERENCE: get_vegas_segment_width returns
  panels, not pixels. CONTRIBUTING: the repo uses flake8/mypy/bandit
  pre-commit hooks, not black/ruff, and tests need requirements-test.txt.
- SKIN_SYSTEM/DEVELOPER_QUICK_REFERENCE: stale module paths.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SXb4mKcAkVaxkeTb3YnAdr

* fix: raise the two new dependency floors past their CVEs, silence a deliberate re-export

All three of these were introduced by this PR, which is what makes them
worth fixing here rather than deferring.

`urllib3` and `jinja2` were added to the requirements so that direct
imports stop relying on transitives — right call, but both floors were
set to the version that introduced the API rather than a version that is
safe to install. `urllib3>=1.26.0` sits below roughly ten CVEs including
a decompression-bomb safeguard bypass, and `jinja2>=3.1.0` below five
including two sandbox breakouts. Raised to 2.7.0 and 3.1.6, which is what
a working device already runs, so no install is disturbed. The comments
now say the floor is a security floor, since the next person to read
"imported directly" would otherwise reasonably lower it again.

This is the same reasoning the PR already applied to werkzeug; these two
just missed it.

The `DateTimeEncoder` import in cache_manager is unused on purpose — the
canonical class moved to src.cache.disk_cache and this re-export keeps
the documented import path working. flake8 cannot see intent, so it gets
an explicit `# noqa: F401` rather than being removed and quietly breaking
anything importing it from here. Verified the re-export still resolves to
the same object and still serialises datetimes.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Udr6MfaFLUPhX5Fgo67Jf5

* fix: address CodeRabbit re-review — installer robustness and doc lint

- install_dependencies_apt.py: an import-only check let Debian
  Bookworm's python3-freetype 2.3.0 satisfy the freetype-py>=2.5.1 pin.
  check_package_installed() now verifies the installed freetype-py
  version, and an apt install that lands below the minimum falls through
  to pip instead of counting as success.
- first_time_install.sh: the .web_deps_installed marker was created even
  when the smart installer failed, so re-runs skipped installation with
  dependencies missing. The marker is now created only on success.
- CONTRIBUTING.md: document installing the pre-commit CLI before
  'pre-commit install' (the requirements files don't provide it).
- Doc lint: fence language on the on-demand cache example (MD040),
  blockquote continuation in GETTING_STARTED (MD028), and the
  suppress_adapter_load_errors key removed from the emulator debug
  example to match the options table.

Skipped one finding with reason (noted on the PR): the per-plugin
display_duration field in PLUGIN_QUICK_REFERENCE's example is not
obsolete — BasePlugin.get_display_duration() reads it and
PLUGIN_CONFIG_CORE_PROPERTIES.md documents it as a core property;
display.display_durations is a per-mode override, not a replacement.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SXb4mKcAkVaxkeTb3YnAdr

---------

Co-authored-by: Claude <noreply@anthropic.com>
 2026-08-06 14:04:23 -04:00
ChuckandClaude Opus 5 f2b246ef03 fix(version): make the core version have exactly one answer (#428) 
* fix(version): make the core version have exactly one answer

Plugin compatibility floors compare against src.__version__, so that string
has to be trustworthy. It has not been. v3.1.0 was tagged 2026-05-31 while
src/__init__.py still said "1.0.0"; the bump did not land until 2026-07-12.
Every device installed from that release reports 1.0.0, which is below the
(2, 0, 0) floor in PluginLoader._warn_if_incompatible -- so those users are
silently exempt from every plugin compatibility warning.

web_interface carried a third answer, a hardcoded "3.0.0" that nothing read
and that had drifted two majors from the core. It now re-exports the
canonical value, so it cannot disagree again.

Adds:

- test/test_version_consistency.py (enrolled in the core unit CI job):
  src.__version__ is parseable semver, matches the newest CHANGELOG heading,
  the CHANGELOG's headings are unique and descending, and web_interface
  tracks the core. src.plugin_system.__version__ is deliberately excluded --
  it versions the plugin API and moves independently.

- scripts/check_release_version.py + a release-version-check workflow that
  asserts the tag, the CHANGELOG and src.__version__ agree. Runs on pushed
  v* tags and published releases, and via workflow_dispatch so a tag can be
  checked *before* it is created:

      python scripts/check_release_version.py v3.2.0

Verified: 757 core unit tests pass including the four new ones; the script
exits 0 for v3.2.0 and non-zero for both a mismatched tag (v3.1.0) and a
non-semver one (v2.5); web_interface and web_interface.app still import.

Prerequisite for cutting v3.2.0 -- phase B4 in docs/SPORTS_UNIFICATION.md.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Udr6MfaFLUPhX5Fgo67Jf5

* fix(version): address review — regex strictness, OSError, stale doc claim

From CodeRabbit on #428, all three valid:

- The module docstring claimed the tag check "runs at release time in
  .github/workflows/release-version-check.yml". That workflow is held back to
  a follow-up PR (the pushing token lacks the `workflow` scope), so the claim
  was false as written. Both files now describe the script as a manual
  pre-flight and say the CI wiring is still to come.

- `\d` also matches non-ASCII decimal digits, which int() happily parses, and
  `\s` matches newlines -- so "##\n3.2.0" read as a version heading. Patterns
  now use [0-9] and [ \t], kept in step across the test and the script, with a
  regression test pinning both behaviours.

- A missing or unreadable CHANGELOG.md raised OSError out of read_text() and
  printed a traceback. In a release gate that reads as "the tooling is
  broken"; it now reports the path and a recovery action and exits 1.

Verified: v3.2.0 passes, a mismatched tag exits 1, and a missing CHANGELOG
exits 1 with the new message instead of a traceback. 5 tests pass.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Udr6MfaFLUPhX5Fgo67Jf5

---------

Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
 2026-08-03 15:21:47 -04:00
ChuckandClaude Opus 5 16fbb7ebeb fix(install): survive the rgbmatrix build on low-memory Pis (#430) 
* fix(install): survive the rgbmatrix build on low-memory Pis

The one-shot installer failed at Step 6 on a 1GB Pi with "Failed building
wheel for rgbmatrix", and told the user to install build tools they already
had. The real cause was the kernel OOM killer.

Upstream's pyproject.toml declares no [tool.scikit-build] options, so
scikit-build-core drives Ninja at its default of nproc+2 jobs -- six
concurrent compiles on a 4-core Pi. CMakeLists.txt compiles the same 14
sources three times (~45 translation units), two of them Cython-generated
C++ where a single cc1plus peaks near 800MB. That does not fit in 512MB-1GB
of RAM.

Add scripts/install/lib_lowmem.sh and wire it into the installer:

- Cap build parallelism at max(1, min(cores, RAM/768)) via
  CMAKE_BUILD_PARALLEL_LEVEL, which is what cmake --build actually reads.
  MAKEFLAGS is ignored by Ninja and is set only as a Makefile-generator
  fallback. A 4GB Pi 4 still gets 4 jobs; 512MB and 1GB boards get 1.
- Add a temporary swapfile sized to bring RAM+swap to 3GB (capped at 2GB),
  removed once the build finishes. An EXIT trap is the backstop for the
  error path. Nothing is written to /etc/fstab or /etc/dphys-swapfile.
  Existing swap is measured excluding zram, which is compressed RAM and so
  does not help a build OOM.
- Keep pip's build tree off tmpfs. Debian 13 mounts /tmp as tmpfs, so the
  default held the whole C++ build tree in RAM alongside the compiler.
- Diagnose OOM failures from the build log and the kernel ring buffer,
  instead of always blaming missing build tools. The OOM killer writes
  nothing to pip's output, which is why this was misreported.
- Report RAM and the chosen job count in the Step 1 preflight, and emit a
  heartbeat during the compile so a deliberately serial 15-25 minute build
  does not look like a hang.

New flags --skip-swap and --build-jobs N, with LEDMATRIX_SKIP_SWAP and
LEDMATRIX_BUILD_JOBS equivalents.

Also skip the duplicate apt-get update that the one-shot installer and
first_time_install.sh each ran a minute apart, and complete the
dphys-swapfile advice in diagnose_dependencies.sh with the CONF_MAXSWAP
line, without which raising CONF_SWAPSIZE above 2048 is silently clamped.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01VdsJs65WnUo8BtKHMAGA1q

* fix(install): address review findings on the low-memory build path

Three fixes from PR review:

- Validate --build-jobs / LEDMATRIX_BUILD_JOBS before check_memory's
  fallback return. When lib_lowmem.sh is absent that return also honoured
  the override, so a non-numeric value skipped validation and instead blew
  up later in an arithmetic test in Step 6 with a generic error.
- Fall back to the default TMPDIR when the disk-backed build directory
  cannot be created, rather than pointing the build at a path that does
  not exist. A nearly-full disk is the likely cause on exactly the devices
  this targets.
- Pass LEDMATRIX_APT_UPDATED explicitly to the sudo child instead of
  relying on -E, which a sudoers env_reset/env_keep policy can strip.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01VdsJs65WnUo8BtKHMAGA1q

* fix(install): don't add 30s to every rgbmatrix build

The build progress heartbeat slept for the full 30s report interval
before re-checking whether the compile had finished, so every build paid
up to 30 seconds of dead wall time -- including fast ones on a Pi 4/5 and
every --force-rebuild run.

Poll every 2s and report every 30s instead. Measured: 30s of overhead on
an instant build drops to 2s, with heartbeats still emitted on the same
schedule.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01VdsJs65WnUo8BtKHMAGA1q

---------

Co-authored-by: Claude <noreply@anthropic.com>
 2026-08-03 15:21:18 -04:00
ChuckandClaude Sonnet 5 5b45f35888 Vegas mode: reclaim dead space and pace the rotation (#423) 
* Vegas mode: reclaim dead space and pace the rotation

On a wide panel Vegas mode spent much of its time showing black. At 50px/s
on a 512px display, one display width of blank is 10.2 seconds, which makes
several long-standing behaviours expensive:

- ScrollHelper prepended a full display width of black as an "initial gap",
  charged once per cycle — 10.2s of black at the start of every rotation.
- Plugins without get_vegas_content() are captured off a full-display canvas,
  so their blank margins entered the ticker too. Measured: of-the-day drew
  35px of "No Data" on a 512px canvas (92% blank), youtube-stats 142px of
  content with 185px of black either side. Only the scroll_helper path had
  any trimming.
- Cycle transitions deliberately pushed a blank frame and then recomposed
  synchronously: 84ms at best, 4.8s at worst, every millisecond of it black.
- buffer_ahead doubled as the cycle size, so a 21-plugin install showed 3
  plugins per cycle and took ~7 cycles to come around.
- separator_width was applied between every image rather than at plugin
  boundaries, so a per-row ticker like the F1 scoreboard (116 images, which
  it renders 4px apart internally) got a 32px chasm between each row — and
  the width budget didn't count those gaps, so the plugin quietly occupied
  far more of the panel than intended.

Changes:

- src/vegas_mode/geometry.py: numpy column-ink primitives shared by the
  trimmer and the audit tool, so the number reported is the number acted on.
  A Python per-column loop over a 17,000px strip is far too slow for the
  render path.
- PluginAdapter trims every content path, not just scroll_helper. Only outer
  edges are cropped: interior blank columns are the plugin's own layout
  (logo left, score right) and closing them would corrupt the design. A
  plugin on a non-black background is inherently unaffected.
- ScrollHelper.create_scrolling_image takes an explicit lead_gap, still
  defaulting to display_width so the many standalone-ticker callers are
  unchanged. Vegas passes lead_in_width (default 0).
- Cycle end holds the last rendered frame instead of blanking, turning the
  recompose into a brief freeze rather than the panel switching off.
- plugins_per_cycle (default 6) is split from buffer_ahead, which goes back
  to being only a prefetch low-water mark.
- max_plugin_width_ratio (default 3x display width) caps one plugin's share
  of a cycle. Overflow is deferred, not discarded: a rotation offset advances
  each fetch so later rows appear on subsequent cycles. Single oversized
  images are cropped at a blank column so the cut misses glyphs.
- Composition groups images by plugin: rows are joined by intra_plugin_gap
  (default 8) and separator_width applies only between plugins. The width
  budget now counts those gaps.
- Plugin data updates no longer run on the Vegas render path.

All new settings are user-configurable in Display -> Vegas Scroll, including
min/max cycle duration and dynamic duration, which previously existed in code
but were reachable only by hand-editing config.json.

Measured with scripts/dev/vegas_audit.py on a 512x64 panel:

  mean ink coverage    42.7% -> 69.4%
  fully blank           5.9% -> 0%
  reads as empty        13.6% -> 0%
  worst blank stretch    4.8s -> 0s
  full rotation          414s -> 123s
  plugins per cycle         3 -> 6

Note the metric choice: a "fully blank" scan (>=95% black viewport) reported
only 0.4% and badly understated the problem, because two full-width segments
with mid-canvas content never fully blank the viewport — they hold it at ~28%.
window_coverage_stats grades every viewport position by how much ink it
carries, which is what tracks perceived dead time.

Known remaining: cycle transitions still freeze ~3.5s while the next cycle is
fetched. Fixing that needs background prefetch, which is deferred because the
fallback-capture path mutates the shared display_manager.image and racing it
against the render loop risks torn frames.

Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01KEZK1P1Q1fu5pcuVrkrCFZ

* Drop unused Optional import from the vegas audit script

Flagged by Codacy (F401). Any, Dict and List are all still used.

Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01KEZK1P1Q1fu5pcuVrkrCFZ

* Align Vegas API bounds with validate(), fix audit config plumbing

Both from review feedback on #423.

The web API's accepted ranges disagreed with VegasModeConfig.validate(),
which is what actually gates Vegas starting:

  scroll_speed      1-100  -> 1-200   (a slider value of 150 returned 400)
  separator_width   0-500  -> 0-128
  target_fps        1-200  -> 30-200
  buffer_ahead      1-20   -> 1-5

The three loose ones were the dangerous direction: the value saved with a
200, then VegasModeCoordinator.start() failed validation with only a log
line, so the ticker silently never ran. The UI already matched validate() in
all four cases, so the API was the odd one out.

test_vegas_api_bounds_match_validate parses the numeric_fields map out of
api_v3 and asserts every bound against validate(), plus that validate()
accepts both endpoints and rejects just outside them, so these cannot drift
apart again. That test immediately caught a missing upper bound on
min_plugin_width, now added — unbounded it would drop every segment and
leave a blank ticker.

Separately, vegas_audit.py constructed PluginAdapter without the config, so
it fell back to VegasModeConfig() defaults and would report trimming and
width-budget behaviour that differed from the user's config.json. It now
passes the loaded config exactly as the coordinator does. This is the same
class of drift the explicit lead_gap and grouping arguments already guard
against. Output is unchanged on a rig whose config matches the defaults.

Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01KEZK1P1Q1fu5pcuVrkrCFZ

* Vegas mode: render plugins narrower, space rows by measured separation

Trimming reclaims blank margins but cannot compact a layout that genuinely
spans the display — a five-column forecast, a progress bar drawn at 100%
width, a stat block with the panel's whole width between its elements. Those
need the plugin to make different layout decisions, which means telling it the
screen is narrower while it renders.

DisplayManager.render_size() presents a smaller logical canvas for the
duration of a Vegas content fetch, reusing the same _LogicalMatrix
indirection double-sided mode already relies on so plugins see a consistent
size from every accessor. Plugins that size themselves from matrix.width need
no changes at all; one that wants to be explicit can read the new
BasePlugin.get_vegas_render_width().

Width is a percentage so a single setting travels across panel sizes:
vegas_scroll.render_width_pct globally, or vegas_width_pct in an individual
plugin's config. Measured on a 512x64 panel with real data:

  ledmatrix-weather   1536px -> 576px   (forecast becomes narrow cards)
  youtube-stats        353px -> 199px   (2% blank left, so genuinely compact)
  geochron             453px -> 153px   (ink density rises to 100%)
  ledmatrix-flights    950px -> 740px

The youtube-stats figure is the clearest evidence the layout itself changed
rather than being cropped: at full width the content had to be trimmed from
512px to 353px, whereas at 40% it arrives with almost no blank to reclaim.

Row spacing is now measured rather than added. A flat gap gets it wrong in
both directions at once — content drawn flush to its own edges ends up nearly
touching (reported for recent sports scores, which sat 8px apart), while
content already carrying wide margins gets pushed even further out.
separation_gap() measures the blank each pair already has and adds only the
shortfall, up to min_content_separation (default 24). intra_plugin_gap stays
as a floor applied regardless.

Two tests shipped in the previous commit encoded the old flat-gap arithmetic
and are updated to the measured semantics, including one renamed to reflect
that zero intra_plugin_gap alone no longer butts rows together.

Also fixes a real bug found while testing: the harness display manager had no
render_size(), and because the adapter catches broadly that surfaced as "no
content" rather than an error, silently dropping five plugins. Added the
context to VisualTestDisplayManager for parity, and _render_at() now degrades
to a no-op on any display manager lacking it, so a third-party or older
harness loses the narrowing rather than the content.

Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01KEZK1P1Q1fu5pcuVrkrCFZ

* Vegas mode: end cycles before the wrap, keep the width budget honest

Three fixes, the first a regression from lead_in_width defaulting to 0.

get_visible_portion wraps: once scroll_position + display_width passes the end
of the strip it fills the right of the frame from the *head* of the same strip.
So the final display_width of travel showed the cycle's first plugin re-entering
on the right while its last plugin exited on the left, and the recompose that
followed replaced both at once. On a 512px panel at 50px/s that was 10.2s of
two plugins on screen at once, ending in a hard cut — reported as the ticker
"switching mid-scroll" from F1 to news.

That used to be invisible because the strip began with a full display_width of
blank, so the wrapped-in region was black. Removing that blank (it was 10s of
dead panel per cycle) exposed the wrap. Cycles now end one display width
earlier, before any wrapped content appears, clamped for strips no wider than
the display so they don't complete instantly and spin the recompose loop.

Verified on hardware: a 3936px strip now completes at 68.5s, exactly
(3936 - 512) / 50.

Second, auto_trim=False also skipped the width budget, which is an unrelated
concern — turning off margin cropping should not let one plugin hold the panel
for minutes. Seen in the field: the F1 scoreboard contributed 116 images and
14,848px untouched, giving a 33,821px cycle (11 minutes of content). The budget
now applies regardless of trimming; with it restored that cycle is 6,362px.

Third, the budget accounted for row gaps using the flat intra_plugin_gap while
the compositor had moved to measured separation, so it under-counted by up to
(min_content_separation - intra_plugin_gap) per row and a many-row plugin
overran its cap. Both now use the same separation_gap() rule, and a test
asserts the composed block fits the budget end to end rather than trusting the
two paths to agree.

Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01KEZK1P1Q1fu5pcuVrkrCFZ

* Fix IndexError in find_blank_cut when the cut lands on the image edge

A cut position after the last column is legitimate — _crop_to_budget asks for
min(start + budget, img.width), which equals the width whenever the remaining
strip is shorter than the budget. find_blank_cut clamped target to width but
then walked leftwards starting at target itself, so ink[width] raised
IndexError.

Caught on hardware: it killed the ledmatrix-stocks fetch, and because
_fetch_plugin_content catches broadly that surfaced as the plugin silently
contributing nothing for the cycle.

Only reachable on the second or later pass of the rotating window over a single
oversized image, which is why the existing tests missed it — they all exercised
the first pass, where start is 0 and start + budget is comfortably inside the
image. Added TestRotationAcrossMultipleCycles, which walks the window round
several times and asserts content is never lost, plus direct coverage of
find_blank_cut at and beyond the image edge.

Both bounds now stop at width - 1 so neither direction can index past the end.

Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01KEZK1P1Q1fu5pcuVrkrCFZ

* Only cut oversized segments at real gaps between items

The width-budget crop snapped to the nearest blank column, and in rendered text
the gap between two characters is a single column. So a cut routinely landed
inside a word: the cycle showed "Wednesda" and the orphaned "y" turned up as a
lone floating letter in the next cycle, positioned after whatever plugin
happened to precede it.

Measured on the clock-simple segment to confirm: its blank runs are
[1, 1, 1, 1, 1, 8, 8] — five single-column letter gaps, every one of which
find_blank_cut would happily have chosen.

Cuts now only land in a run of at least min_cut_gap blank columns (default 6),
which excludes letter spacing while still finding the gaps plugins put between
items (the stocks ticker uses 32px, baseball 48px). Where no boundary falls
inside the budget the cut waits for the next one and overruns, because
splitting an item is worse than a slightly long segment.

Continuous content is treated differently on purpose: an image with no internal
gaps is a map or a chart, where any column is as good as another, so it is still
cut to the budget exactly. The gap rule protects discrete items; letting a solid
image escape the cap in its name would be wrong.

blank_runs() is vectorised — 48ms for a 17,000px strip, against seconds for a
per-column Python loop.

Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01KEZK1P1Q1fu5pcuVrkrCFZ

* Hold capture_mode for every plugin render, not just narrowed ones

The native content path only entered capture_mode when it was also narrowing
the canvas, so at full width — which is every plugin without a vegas_width_pct
override, i.e. most of them — a plugin calling update_display() while building
its Vegas content wrote straight to the hardware. That is a visible flash
mid-scroll, and it lines up with the flash reported at cycle transitions, when
several plugins are fetched back to back.

Suppression is now unconditional; the narrowing context stays separate because
it is already a no-op at full width.

Both contexts are reached through helpers that degrade to nullcontext when the
display manager lacks them. That matters more than it looks: the adapter's
handlers are deliberately broad, so an AttributeError from a missing context
does not surface as an error — it surfaces as the plugin contributing nothing.
Making the call unconditional without this turned 44 tests red for exactly that
reason, all of them reporting lost content rather than the real cause.

The test double now provides capture_mode and render_size too, so tests
exercise the real contexts instead of silently taking the degraded path.

Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01KEZK1P1Q1fu5pcuVrkrCFZ

* Vegas mode: one continuous strip instead of swapping cycles

A cycle used to be a discrete strip that got replaced: motion stopped, every
pixel was substituted at once, and the next group started with the viewport
already full. That is the freeze, the flash and the jump.

The strip is now extended rather than replaced. ScrollHelper gains
append_content(), which adds items on the right without touching
scroll_position or total_distance_scrolled, so motion continues and the next
group simply arrives from the right. Because completion is measured against
total_scroll_width, extending also defers completion — there is no longer a
cycle boundary to see.

drop_scrolled_prefix() reclaims what has gone past, keeping the strip bounded
however long Vegas runs (observed 5,000-11,000px against an unbounded strip
otherwise). It shifts total_distance_scrolled and total_scroll_width together so
the completion arithmetic is unchanged, and refuses to run while the viewport is
wrapping: wrapping reads the head of the strip into the right of the frame, so
trimming the head there would visibly change the picture. A test caught that.

Groups are prepared off the render thread. The constraint is that the canvas and
the matrix proxy are process-wide mutable state, so narrowing or capturing
through them from another thread would corrupt the frame the render loop is
pushing. get_content() therefore takes offscreen_only: the background thread uses
only paths that avoid the canvas, and anything needing it is marked and picked up
on the render thread. That puts the expensive work (native renders of leaderboard
and baseball cards, seconds each) in the background and leaves the cheap work
(display capture, 40-600ms) in the foreground.

DisplayManager's capture flag is now thread-local. As a shared flag, a background
capture would have suppressed the render loop's own frame pushes for its
duration, freezing the panel precisely when the point was to avoid a freeze.

Canvas-bound plugins are drained one at a time rather than as a batch: six at
once held the render thread for 1.75s. Drains are also spaced by two seconds
while the lookahead is healthy, since taking them back to back turns one long
stall into a run of short ones. When the strip is genuinely running short the
throttle is ignored, because content matters more than smoothness there.

Measured on hardware: zero cycle-complete swaps, drains landing 2-4s apart,
lookahead holding at 1,200-3,500px, no errors.

Set continuous_scroll false to restore the swap behaviour; the old path is intact.

Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01KEZK1P1Q1fu5pcuVrkrCFZ

* Pace the Vegas frame loop adaptively: 31.5 -> 78.7 fps

The loop slept a fixed frame_interval on top of however long the frame took, so
at a measured 31.6ms per frame a flat 8ms of that was pure idle — a quarter of
the budget spent not rendering. It now sleeps only the remainder of the budget.

Measured on hardware: 31.5 fps to 78.7 fps sustained, with CPU going *down* from
150% to 127%. Scroll speed is unchanged at 49.9px/s against a configured 50,
because motion is derived from elapsed time rather than frame count — this buys
smoothness, not speed.

Worth recording what the bottleneck was not: the per-frame render path measures
0.34ms in total (0.18ms for the numpy slice, 0.17ms for the dirty-tracking
digest), which is a theoretical 2900 fps. Optimising any of that would have been
wasted effort. The frame was idle, not busy.

Also nices the prefetch thread. Its work is PIL and numpy that releases the GIL,
so the scheduler can act on the priority, and without it the prefetch competes
for the same cores as the render loop.

Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01KEZK1P1Q1fu5pcuVrkrCFZ

* Sub-pixel scrolling: motion at the frame rate, not the pixel rate

With integer positioning the number of distinct frames per second equals the
scroll speed in px/s, however fast the loop renders. Measured at 50px/s and
78.7fps, 36% of frames were byte-identical: the extra frames cost work and
bought no motion, and what was left was 50 discrete 1px steps a second.

Two things were wrong with the pre-existing sub-pixel support. get_visible_portion
never consulted sub_pixel_scrolling — it always took the integer path, so the flag
and _get_visible_portion_subpixel were dead code. And that implementation needed
scipy.ndimage.shift, which is not installed on the target devices (HAS_SCIPY is
False there), so it would not have interpolated even if reached. Verified both:
positions 1000.0 and 1000.5 produced identical frames either way.

Blending is now wired up and implemented with numpy. Two details make it
affordable: slice cached_array directly instead of building two PIL images only
to convert them straight back (the naive version measured 15x the integer path),
and use fixed-point uint16 multiply-add rather than float32, which suits the Pi's
cores and gives finer weighting than the panel can resolve. Result 0.939ms
against 0.237ms — 0.70ms added per frame, a 1065fps ceiling.

Measured on hardware: 81.2 fps with blending on, against 78.7 with it off, so no
cost within noise — and every frame is now a distinct position rather than one in
three being a repeat.

The trade is a slight horizontal softening of text, since each frame blends two
positions. Set smooth_scroll false for maximum crispness.

Also benchmarked and cleared as non-issues: extending the strip costs 9.4ms on an
11,000px strip and trimming 2.5ms, both under one frame at this rate.

Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01KEZK1P1Q1fu5pcuVrkrCFZ

* Add overflow handling: keep ordered content whole instead of rotating a window

The width budget split any oversized plugin by advancing a window each cycle.
That is right for interchangeable items — news headlines, odds, stock prices —
but wrong for ordered content: a league table showed ranks 1-6, then resumed at
7 two rotations later, which reads as out of order and out of context. Nobody
needs rank 23 in a ticker; they need the top of the table, every time.

overflow_mode chooses between them:

  rotate   — advance a window each cycle so everything is seen eventually
             (unchanged default)
  truncate — always show the start and drop the rest, keeping ordered content
             coherent. Records no window state, so every pass starts at the top.

Per-plugin vegas_overflow overrides the global setting, since one install has
both kinds of plugin. Also adds per-plugin vegas_max_width_screens, so content
that must stay whole can be given more room — or uncapped with 0 — without
lifting the cap on every ticker.

Applied on the test rig: f1-scoreboard and ledmatrix-leaderboard set to
truncate, and baseball given 4.5 screens because it was showing 8 of 9 games
when the whole slate needed only a little more room. Verified: F1 now reports
"the first 10 of 116 ... the rest are not shown", baseball has dropped out of
the budget log entirely, and stocks, odds-ticker and stock-news still rotate.

Also corrects the crop log, which claimed "window advances next cycle"
unconditionally and so misreported truncated crops. A test now pins the
behaviour behind the message: truncate must leave no offset recorded.

Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01KEZK1P1Q1fu5pcuVrkrCFZ

* Stop Vegas mode showing last night's games as if they were live

A game that was live in the evening was still being drawn as live the next
morning. Two faults combined to freeze plugin visuals indefinitely.

PR #291 added a call to plugin_adapter.invalidate_plugin_scroll_cache() so
a plugin's own cached scroll image would be rebuilt from fresh data. That
method was never implemented. hot_swap_content() wraps the call in a broad
except, so every hot swap has raised AttributeError and been swallowed
silently ever since — which is why the visuals it was meant to keep fresh
never were.

Continuous scrolling then removed the only path that reached it at all:
should_recompose() and hot_swap_content() are called from the
non-continuous branch of run_frame(), and continuous_scroll defaults to
True. So on a default install the pending-update flags were set by the
update tick, never consumed, and grew without bound.

Together these froze content completely, because refetching is not enough
on its own: the sports plugins' get_vegas_content() regenerates only "if
the cache is empty", so take_next_group() kept receiving the same picture
however often it asked.

Fixed by:

- Implementing invalidate_plugin_scroll_cache(). It covers both layouts —
  a helper directly on the plugin (stocks, news, odds-ticker) and one
  owned by a scroll-display manager (the sports scoreboards, which is the
  shape that produced this bug) — and clears cached_image and
  cached_array together, since the array is the image's numpy mirror.

- Adding StreamManager.invalidate_pending_updates() and calling it from
  the continuous branch. It only drops the caches; the plugin recomposes
  when it next comes round in the rotation. process_updates() is wrong
  here: it refetches synchronously and merges into the active buffer that
  continuous mode bypasses, and hot_swap_content() rebuilds and
  repositions the whole strip, which is the freeze-and-jump this mode
  exists to avoid.

Tests assert the fix rather than the implementation: 14 of the 17 new
tests fail without it. Includes the wiring itself, since the regression
was a call that was simply absent, and a check that the scroll position is
untouched so this cannot regress into the swap's visible jump.

All Vegas suites pass (355 tests). test_display_controller_vegas_tick.py
still cannot be collected off-device for want of rgbmatrix, identically
with and without this change.

Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01KEZK1P1Q1fu5pcuVrkrCFZ

* Fix two CodeRabbit-flagged test assertions in vegas density tests

test_prepared_group_is_used_without_refetching had a tautological final
assertion; now checks stream.calls directly. test_no_partial_letter_at_either_edge
required both crop edges to be blank, but the left edge here is always the
crop's start position with no lead-in gap in word_strip, so it legitimately
carries ink — only the right edge is an actual cut and needs the check.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01KEZK1P1Q1fu5pcuVrkrCFZ

---------

Co-authored-by: Claude <noreply@anthropic.com>
 2026-07-31 09:40:38 -04:00
ChuckandClaude Fable 5 cdf03fb107 Add skin system: user-installable visual overlays for sports scoreboards (#419) 
* Add skin system: user-installable visual overlays for sports scoreboards

Skins restyle a scoreboard's live/recent/upcoming rendering while the
host plugin keeps doing data fetching, scheduling, caching, live
priority, and vegas mode — the anti-fork alternative for users who only
want a different layout.

- src/skin_system/: ScoreboardSkin API, SkinContext (canvas + adaptive
  layout + logo/font helpers), discovery/loading runtime with API major
  version gating and per-skin module namespacing
- src/base_classes/sports.py: _render_game() seam at the three
  _draw_scorebug_layout call sites; skin-first with built-in fallback,
  3-strikes session disable, slow-render warning; per-mode skin config
- scripts/validate_skin.py: headless multi-mode/multi-size validator
  with bundled per-sport fixtures (no hardware or network needed)
- skins/example-classic-baseball/: working reference skin
- Web UI: served-schema Visual Skin dropdown (validation never
  enum-restricted, so uninstalled skins can't invalidate configs) and
  GET /api/v3/skins
- Store: registry entries with type "skin" install to skins/
- docs/SKIN_SYSTEM.md (architecture), docs/CREATING_SKINS.md (author
  guide incl. Claude Code prompt), view-model contract locked by tests

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01LrCusPasy1qeUN5anK3aA1

* Address review feedback on skin system

- skin_runtime: cache the entry module so the 2nd/3rd load of the same
  skin (live/recent/upcoming hosts) doesn't re-execute it with unbound
  sibling aliases; rebind cached sibling modules to their bare names
  around entry import and restore prior bindings after; include per-
  manifest mtimes in the discovery cache fingerprint so in-place skin
  updates are picked up
- sports.py: count render_skin_card exceptions toward the 3-strike
  session disable
- store_manager: validate skin ids (pattern + resolved-path containment
  in skins/), reject registry/manifest id mismatches, and stage+validate
  downloads in a temp sibling before replacing an existing skin
- schema_manager: leave the schema untouched when the configured skin
  value is a per-mode mapping (a string dropdown could overwrite it)
- validate_skin.py: reject non-positive sizes and non-object --options
  at parse time; support --output-dir outside the repo; type annotations
- example skin: validate accent_color once at load with logged fallback
- fixtures: pregame 0-0 scores in football/hockey upcoming fixtures
- api /skins: rely on the self-invalidating discovery cache instead of
  force_refresh
- docs: valid JSON manifest example, load_logo caching semantics spelled
  out, language ids on fenced blocks
- tests: view-model contract test now exercises the real extractor;
  regression test for repeated same-skin loads with sibling modules

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01LrCusPasy1qeUN5anK3aA1

---------

Co-authored-by: Claude <noreply@anthropic.com>
 2026-07-18 11:07:08 -04:00
ChuckandClaude Sonnet 5 66f9950a30 chore(ci): add security-audit tooling scripts (workflow files pushed separately) (#414) 
* chore(ci): add security-audit workflow and plugin security-proof scripts

- scripts/prove_security.py, audit_plugins.py, generate_report.py --
  automated checks (dangerous eval()/exec() calls, dependency scanning,
  report generation) for plugins.
- .github/workflows/security-audit.yml + bandit.yaml -- CI wiring for
  the above plus gitleaks secret scanning and bandit static analysis.
- .github/workflows/tests.yml -- pytest matrix across Python 3.10-3.12.

Also fixes two Codacy findings while these files are freshly landing:
- prove_security.py: dropped a pointless f-string prefix with no
  placeholders.
- security-audit.yml: pinned gitleaks/gitleaks-action to a full commit
  SHA (matching this repo's existing pinning convention in test.yml)
  instead of the floating v2 tag.

Split out of the original chore/dead-code-removal commit, which had
accidentally bundled this in alongside unrelated dead-code deletions.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01KEZK1P1Q1fu5pcuVrkrCFZ

* chore: drop workflow files -- pushed separately (needs workflow OAuth scope)

* fix(security-tooling): address PR review findings across bandit.yaml, audit_plugins.py, generate_report.py, prove_security.py

bandit.yaml:
- Removed scripts/prove_security.py's file-level exclusion. Ran bandit
  directly to get ground truth: the real false positive is B105 (dict key
  "PASS" misread as password-like), not the eval/exec pattern the old
  comment claimed. Added a targeted # nosec B105 there, and found+fixed
  the identical pattern already present in generate_report.py.
- Left the repo-wide B607 skip as-is: confirmed via AST scan that properly
  narrowing it touches 100+ bare-name subprocess call sites across
  wifi_manager.py, store_manager.py, permission_utils.py, app.py, and
  start.py -- none of which are part of this PR. Out of proportion to fix
  here; flagged as a dedicated follow-up.

scripts/audit_plugins.py:
- SyntaxError/OSError while scanning a plugin file now report CRITICAL
  (blocking) instead of WARNING/INFO -- a file that couldn't be parsed or
  read was never actually checked for danger, so it must not silently
  pass the audit.
- --plugin <name> now tracks whether the requested plugin was found across
  all PLUGIN_BASE_DIRS and exits 1 with a clear error if not, instead of
  silently scanning zero plugins and reporting success.
- The AST visitor now tracks import aliases (import subprocess as sp;
  from builtins import eval as e) and resolves them before checking
  against dangerous APIs, closing a straightforward evasion of every
  PLUGIN-001 through PLUGIN-005 check. Verified against both aliased and
  unaliased evasion patterns.

scripts/generate_report.py:
- _md_table_row now escapes pipe characters and normalizes newlines in
  every cell, so scanner-controlled content (a matched secret, a bandit
  issue_text) can't corrupt the Markdown table structure.
- _load now distinguishes "artifact missing/malformed" from "valid empty
  result": each summarizer returns an availability flag, and main() now
  reports INCOMPLETE (not PASSED) with exit code 1 when any artifact is
  unavailable, instead of silently folding it in as 0 findings.
- Gitleaks suppression now uses exact-match placeholder values (pulled
  from the actual config_secrets.template.json) plus a template-path
  allowlist, replacing broad substring checks that could hide a real
  secret containing something like "example.com" as part of its value.

scripts/prove_security.py:
- T1b (dangerous plugin calls): a file that fails to parse/read now
  reports CRITICAL with the exception details instead of being silently
  swallowed by `except (SyntaxError, OSError): pass`.
- T6 (Docker hardening): base images must now be pinned to an @sha256
  digest; a specific tag like python:3.12 is mutable and is now correctly
  flagged as unpinned, not just missing tags or :latest.
- T2a (API surface): no config mechanism for enforcing local-only access
  exists in this codebase today (app.py hardcodes host='0.0.0.0'), so the
  "environment-aware" check as described isn't buildable without adding
  new config infrastructure -- out of scope here. Applied the achievable
  part: upgraded from INFO to WARNING, since enforcement can never
  currently be confirmed.
- T1a (zip-slip): replaced the whole-file substring check with an AST
  walk that finds every extract()/extractall() call and confirms an
  is_relative_to() guard + "Zip-slip detected" log precede it in the same
  function. Verified it still passes on the real store_manager.py (both
  the per-member and validate-then-bulk-extract call sites) and correctly
  flags a synthetic unguarded extractall().
- T3a (hardcoded secrets): violation details no longer include the
  matched credential text -- only file, line, pattern type, and a
  redacted SHA-256 fingerprint, so a real finding doesn't get published
  into CI logs/artifacts/PR comments with wider exposure than the
  original leak. Verified with a synthetic secret that no raw content
  reaches the output.

Validated: all four files compile; bandit scans all three scripts clean
(2 legitimate targeted suppressions, 0 unaddressed findings); each new/
changed code path exercised directly (alias evasion, unmatched --plugin,
missing/malformed/valid-empty artifacts, digest-pinning, zip-slip
guard/no-guard, secret redaction); full audit_plugins.py -> prove_security.py
-> generate_report.py pipeline run end-to-end producing a correct report.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01KEZK1P1Q1fu5pcuVrkrCFZ

* fix(security-tooling): address follow-up review findings on PR #414

scripts/generate_report.py:
- Added an explicit `object` type annotation to _md_sanitize_cell's value
  parameter -- it deliberately accepts any stringifiable value (calls
  str(value) unconditionally), so `object` reflects its actual contract
  more accurately than leaving it untyped.

scripts/prove_security.py:
- Dockerfile FROM-line parsing: renamed the comprehension variable `l` to
  `line` (ambiguous single-letter name). More importantly, fixed a real
  false-positive: `FROM --platform=<platform> <image>` was reading the
  --platform= flag itself as the image token, so a properly digest-pinned
  image behind a platform flag was incorrectly reported as unpinned.
  Verified against platform+digest, platform+tag-only, and digest+AS-alias
  Dockerfiles.

scripts/audit_plugins.py:
- Consolidated visit_Call's dangerous-API detection: previously, alias
  resolution only covered ast.Name calls for eval/exec/compile and
  ast.Attribute calls for subprocess/os.system, missing from-imported
  subprocess/os functions called as bare names (from subprocess import
  run as prun; prun(cmd, shell=True) or from os import system as s;
  s(cmd)). Added _resolve_call_target() to resolve both call shapes to a
  single fully-qualified target, then run all five PLUGIN-00x checks
  against that one resolved value. Verified against 10 evasion
  combinations (from-import aliases, direct/attribute calls, aliased
  module imports) and confirmed zero false positives on benign os/
  subprocess usage without shell=True.

Validated: all three files compile, bandit scans clean (same 2 legitimate
suppressions as before, 0 new findings), audit_plugins.py/prove_security.py
re-run against the real repo with no regressions from the prior fix pass.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01KEZK1P1Q1fu5pcuVrkrCFZ

---------

Co-authored-by: Claude Sonnet 5 <noreply@anthropic.com>
 2026-07-15 14:33:17 -04:00
ChuckClaude Sonnet 5Chuck
 		2a1c47fa76 chore: remove dead modules and unused dependencies (~1,180 LOC) (#412) 
* chore: remove dead modules and unused dependencies (~1,180 LOC)

Deletions, each re-verified with a fresh repo-wide grep (core, web,
scripts, docs, plugin monorepo) immediately before removal:

Modules with zero live importers:
- src/background_cache_mixin.py + src/generic_cache_mixin.py (134+150
  LOC — referenced only by each other)
- src/font_test_manager.py (134 LOC)
- src/image_utils.py (22 LOC, self-documented deprecated)
- src/layout_manager.py (408 LOC — only its own test imported it) +
  test/test_layout_manager.py
- src/common/basketball_plugin_example.py (328 LOC sample)

requirements.txt entries with zero importers in core (pre-plugin-era
manager deps): icalevents, geopy, timezonefinder, unidecode. Plus the
google-auth trio (google-auth-oauthlib, google-auth-httplib2,
google-api-python-client): their only importer is the calendar PLUGIN,
which declares all three in its own requirements.txt (verified in the
monorepo and on an installed copy) — the plugin dependency installer
owns them. Existing venvs are unaffected (removal doesn't uninstall);
fresh installs get them when calendar is installed.

Two stale references cleaned (a comment in test_pillow_compat.py, a
directory listing in HOW_TO_RUN_TESTS.md). Full suite green except the
two documented pre-existing failures (circuit_breaker mock drift, fixed
in #400; clock-simple 64x32 overflow, pre-dates this series); all core
entry modules verified importing cleanly under the emulator.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FqzC1nzTWL4kaqgMaQZFam

* fix: remove content accidentally bundled into the dead-code-removal commit

The previous commit's git add/commit swept in a lot of unrelated,
unreviewed work alongside the intended dead-code deletions: a new Plugin
Composer web UI, new security-audit/CI tooling, a new march-madness
plugin, and 23 local-development-only symlinks under plugin-repos/ (per
scripts/setup_plugin_repos.py's own docstring, these are meant to be
generated locally, never committed -- .gitignore has no entry for them,
which is how they slipped in).

Removed here, split into their own PRs instead (except plugin-repos/*
symlinks and march-madness/ncaa_logos, which are dropped rather than
carried forward -- see PR discussion):
- web_interface/blueprints/composer.py + composer-app.js +
  composer-canvas.js + composer.html + manager.py.j2
- scripts/prove_security.py, audit_plugins.py, generate_report.py
- .github/workflows/security-audit.yml, .github/workflows/tests.yml,
  bandit.yaml
- All plugin-repos/* symlinks (local dev artifacts, not meant to be
  committed at all)
- plugin-repos/march-madness/* and the 4 new assets/sports/ncaa_logos/*
  PNGs it needed (left out of every split PR pending a decision on
  whether march-madness belongs in the core repo or the plugin monorepo)

This PR now contains only what its title describes: the dead-code
removal from the previous commit, untouched.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01KEZK1P1Q1fu5pcuVrkrCFZ

---------

Co-authored-by: Chuck <chuck@example.com>
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
 2026-07-15 09:58:31 -04:00
ChuckClaude Sonnet 5Chuck
 		14a59c863c perf(wifi): one status fetch per monitor tick instead of three (#411) 
* perf(wifi): one status fetch per monitor tick instead of three

The wifi monitor daemon fetched WiFi status + ethernet state before its
AP-mode check, the check internally fetched the same state again (with
retry), and the daemon fetched a third time afterwards — each fetch is
several nmcli subprocess forks, every 30s, forever, even on a perfectly
healthy link.

check_and_manage_ap_mode's decision logic is extracted to
_manage_ap_mode(status, ethernet, ap_active); the new
check_and_manage_ap_mode_with_state() runs the single (retrying) fetch
battery and returns (changed, status, ethernet, ap_active_after) — the
post-state is derivable because state only ever flips via one enable or
one disable. The original bool-returning method delegates, so existing
callers are untouched. The daemon's dead pre-fetch is removed and its
post-check reads use the returned state; the AP-enable retry semantics
(_get_wifi_status_with_retry) are preserved exactly. ~10-15 forks/30s
drops to ~4-6.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FqzC1nzTWL4kaqgMaQZFam

* fix(wifi): add missing type hints on AP-mode state helpers, fix unused-var lint in tests

- check_and_manage_ap_mode_with_state now declares its
  Tuple[bool, WiFiStatus, bool, bool] return type instead of being untyped.
- _manage_ap_mode's status/ethernet_connected/ap_active parameters are now
  typed (WiFiStatus, bool, bool), matching its existing -> bool return hint.
- test_wifi_check_state.py: rename unused unpacked variables (status/
  ethernet/ap_after) to underscore-prefixed names at the three call sites
  that don't use them, leaving the used ones untouched.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01KEZK1P1Q1fu5pcuVrkrCFZ

---------

Co-authored-by: Chuck <chuck@example.com>
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
 2026-07-14 11:22:37 -04:00
Chuck 3d347a368a fix(testing): stop plugin enabled:false schema defaults from silently disabling harness tests (#408) 
check_plugin.py, render_plugin.py, and the pytest plugin matrix each built
config as {"enabled": True} then merged in config_schema.json's defaults on
top, letting a plugin's own enabled:false default (a reasonable choice for
a seasonal/opt-in plugin -- 15 of 23 real plugins ship one) silently win.
Every harness/CI render of those plugins was testing "disabled, do
nothing" rather than real behavior.

Extract build_full_config() into testing/loading.py (already the shared
home for plugin-discovery/config-default logic) and use it from all three
call sites: schema defaults, then a forced enabled=True, then harness.json's
config, then the caller's explicit config -- so a test can still
deliberately disable a plugin on purpose, it just can't happen by accident
via the plugin's own shipped schema default anymore.
 2026-07-14 08:21:35 -04:00
ChuckClaude Fable 5Chuck
 		7f7f0d6464 feat: adaptive layout system — size-aware regions, crisp font ladders, image fitting (#393) 
* feat(layout): adaptive layout & font scaling system for plugins

Add src/adaptive_layout.py — opt-in core helpers so plugins render
legibly on any panel size without hand-tuned per-display layouts:

- Region: integer rect algebra (bands/columns/weighted splits/centering)
  that partitions space so text bands can't overlap by construction
- Font ladders: ordered (family, size) steps known to render crisply
  (LADDER_GRID: X11 BDFs at native sizes; LADDER_ARCADE: PressStart2P at
  8px multiples) — fitting walks the ladder instead of scaling pixel
  fonts fractionally
- LayoutContext: breakpoint tiers, geometry scale vs. a declared design
  size, and cached fit_text/fit_lines/font_for_rows queries

Generalizes the three patterns proven in the field: f1-scoreboard's
scale factor, masters-tournament's tiers, baseball-scoreboard's font
fallback ladder.

Wiring: BasePlugin gains a lazy .layout property and draw_fit();
FontManager gains get_native_bdf_size() and a cache_generation counter;
manifest schema gains display.design_size and requires.display_size
max_width/max_height; 96x48 joins DEFAULT_TEST_SIZES; the bounds-check
harness records negative-coordinate draws; TextHelper's broken
measurement helpers are fixed.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* feat(layout): adaptive image fitting + composite region helpers

Add src/adaptive_images.py — the image counterpart to fit_text:
- fit_image(img, box, mode=contain|cover|fill_height|stretch,
  crop_to_ink, anchor, resample, upscale) promoting the proven plugin
  patterns (football's crop-to-ink fill-height logos, masters' cover
  crop + NEAREST flags, static-image's letterbox). Upscales by default —
  thumbnail()'s downscale-only behavior is why imagery stays tiny on
  big panels.
- draw_fitted_image() pastes aligned within a Region with alpha mask.
- One central Pillow>=9.1 RESAMPLE shim replacing ~15 plugin copies.

LayoutContext.fit_image() caches results per (identity, box size,
options) with a 64-entry LRU; id()-keyed entries pin the source image.
BasePlugin.draw_image() is the one-liner adoption path beside draw_fit.

Composites in adaptive_layout.py: Region.offset() (user x/y-offset
passthrough), scoreboard_regions() (the two-logos-plus-score card math
duplicated across six sports plugins, logo_slot = min(H, W//2)), and
media_row() (art-left/text-right).

Fix LogoHelper's size-blind cache key (stale sizes on panel change);
deprecation note on dead image_utils.py.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* feat(harness): scale-up fill check, config variants, multi-size dev gallery

Quality gates for adaptive layout:

- fill_metrics()/check_scale_up() in the safety harness: overflow catches
  content too big for a panel, but nothing caught content that stays tiny
  on panels >= 2x the plugin's declared design size. The check measures
  lit-content extents and warns (or fails, when a plugin opts into
  "fill_check": "strict" in test/harness.json) below 50% coverage on the
  doubled axis. Warn-only by default so no existing plugin breaks.

- harness.json "variants": extra runs with config overlays and their own
  golden dirs, so an opt-in mode (e.g. layout_mode: adaptive) is golden-
  tested beside the classic default. check_plugin.py loops base + variants
  and labels variant results mode@name.

- Dev preview server: GET /api/sizes (harness size sample), POST
  /api/render-matrix (render at up to 12 sizes in one call), size-preset
  dropdown, and an "All Sizes" side-by-side gallery in the preview UI.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* feat(plugins): adaptive-lib discoverability + advisory version compat warning

Discoverability: re-export the adaptive layout/image API from src.common
(the blessed-helpers package plugin authors already know) — canonical
paths stay src.adaptive_layout / src.adaptive_images so nothing breaks.
Document it in src/common/README.md and cross-link ADAPTIVE_LAYOUT.md
from the developer docs authors actually read (quick reference, API
reference, advanced dev, font manager, dev preview, plugin dev guide);
ADAPTIVE_LAYOUT.md gains adaptive-images, composite-layouts and
preserving-user-customization sections.

Compat: PluginLoader now logs one advisory warning (never raises) when a
plugin's manifest declares a min LEDMatrix version newer than the running
core, checking the min_ledmatrix_version / requires.* / versions[]
spellings found in the wild. Guarded against stale core version numbers.

src/__init__.py __version__ bumped 1.0.0 -> 3.1.0 to match the latest
release tag (v3.1.0) — it had never been updated and the compat check
needs a truthful number. NOTE: verify this matches the intended release
numbering before the next tag.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* feat(layout): add measure_font_crispness — verify a ladder rung isn't blurry

PIL antialiases TTF outlines by default; a 'pixel-style' font only
rasterizes without antialiasing at specific sizes (for PressStart2P:
exact multiples of its 8px design grid). A ladder rung at an unverified
size silently renders blurry on an LED panel — this exact bug shipped in
both text-display's and football-scoreboard's custom TTF ladders
(non-8-multiple PressStart2P sizes, and '5by7.regular'/'4x6-font' at
sizes that were never actually crisp).

measure_font_crispness(font, sample_text) renders the sample and reports
the fraction of ink-bbox pixels that are neither pure black nor pure
white. BDF fonts (real bitmaps) always score 0.0; TTF ladders should be
verified against this before shipping — see the new
TestFontFitting::test_ladder_arcade_is_crisp pattern.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* feat(layout): add fit_text_proportional — proportional sizing vs. always-maximize

fit_text always picks the largest ladder rung that fits its box. That's
right when an element owns dedicated space, but wrong when several
independently-fitted elements need to stay visually harmonious as the
panel grows: a score's box might have generous room while a neighboring
logo scales by a fixed geometry factor via px() — fit_text lets the score
balloon out of proportion (even overlapping the logo) even though its
individual pick is technically correct.

fit_text_proportional(text, box, base_size_px, ladder) instead targets
base_size_px * self.scale (the same scale factor px() already uses),
picking the nearest ladder rung at or below that target, still capped to
what fits the box, floored at the smallest rung when the target is below
every rung. Refactored the shared largest-that-fits/ellipsize walk into
_walk_ladder() so fit_text and fit_text_proportional don't duplicate it.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* feat(layout): fit_text_proportional gains an axis-specific scale override

self.scale (min(width_ratio, height_ratio)) is the right conservative
default for anything whose aspect ratio matters, but a caller whose
surrounding composition already scales along a single axis — e.g.
football-scoreboard's logo_slot = min(height, width // 2), which tracks
height alone — needs text sized the same way, or it reads as
under-scaled next to logos that grew on a panel that only got taller
(128x32 -> 128x64: self.scale stays 1.0 since width didn't grow, but
logos still double).

fit_text_proportional(..., scale=None) now accepts an explicit override;
None keeps the existing self.scale default.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(layout): scoreboard_regions reserves real center space at 2:1 aspect ratios

logo_slot = min(height, width // 2) has a blind spot: at exactly 2:1
aspect ratio (width == 2 * height -- a very common shape: two, four, or
more square modules stacked into a taller panel) width // 2 and height
are equal, so the two logo slots claim the ENTIRE width and leave zero
pixels for a center column, no matter how large the panel gets. Not a
'small panel' problem -- 96x48, 128x64, and 256x128 (all exactly 2:1) hit
it identically, while the 128x32 design baseline and panels like 192x48
or 256x32 never do, because height is already the tighter constraint
there.

Two new parameters fix it in the one shared helper every scoreboard-style
plugin composes through:

- min_center_fraction / min_center_design_px reserve at least
  max(width * fraction, design_px * ctx.scale) for the center column,
  capping logo_slot further when needed. The scaled design-px term
  matters on small panels where a flat fraction alone reserves too little
  absolute space.
- score_bleed_fraction extends the score's own fit box (not the logo
  slots themselves) a controlled amount into each side -- the same way
  real broadcast scoreboards let a big score number's edges cross into
  the team marks flanking it. Without this the reserve alone can still be
  too narrow for a short score to render without truncating.

score_area is now genuinely narrower than the full card width (previously
identical to status_band/detail_band, which still span the full width and
overlay the logos -- short text there was never the problem).

Verified against the full harness size spread: a real game score like
'17-21' never needs ellipsis at any tested 2:1-or-tighter aspect ratio
(test_score_never_needs_ellipsis_for_a_short_score), and wide panels
(128x32/192x48/256x32-style) are provably unaffected.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* docs: document scoreboard_regions' center-reserve and score-bleed params

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix: address CodeRabbit review on PR #393

- docs: scope the self.layout note to BasePlugin subclasses (others build
  a LayoutContext directly) and make explicit that adaptive layout is
  opt-in — classic rendering stays unless a plugin adopts the APIs.
- dev_server: broaden the render-request catch (a bad manifest.json now
  returns a clean 400 instead of an unhandled 500) and stop echoing raw
  exception text in the loader-failure responses — full tracebacks go to
  the dev server's console log instead.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FqzC1nzTWL4kaqgMaQZFam

* fix(dev-server): allowlist plugin_id before any path lookup

CodeQL (py/path-injection): plugin_id arrives in request input and flows
into filesystem paths via find_plugin_dir. Gate it with the same
^[a-zA-Z0-9_-]{1,64}$ allowlist the web UI's pages_v3 uses, at the
single choke point every route resolves through.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FqzC1nzTWL4kaqgMaQZFam

* fix(dev-server): lexical containment check on resolved plugin dirs

CodeQL doesn't recognize the interprocedural allowlist as a
path-injection barrier; add the canonical one — normalize (without
following symlinks, since dev plugins are commonly symlinked into
plugins/) and require the result to stay inside the search dir.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FqzC1nzTWL4kaqgMaQZFam

* fix(dev-server): inline normpath containment barrier before render

CodeQL doesn't credit the sanitization inside find_plugin_dir along
this flow; apply its documented barrier (normpath + startswith against
the allowed roots) inline in _parse_render_request, on the exact path
that reaches the render/load sinks.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FqzC1nzTWL4kaqgMaQZFam

* fix(dev-server): derive plugin dir from trusted directory listings

CodeQL's barrier-guard recognition doesn't see a startswith check
inside an any() comprehension, so the normalize-and-prefix approach
still flagged. Break the taint outright instead: after lookup, re-derive
the directory by enumerating the search dirs (iterdir) and matching by
path equality — the Path used for all downstream file access is built
solely from trusted listings, never from request input.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FqzC1nzTWL4kaqgMaQZFam

* fix(dev-server): use os.scandir for path-injection barrier, redact stack traces from render responses

CodeQL doesn't model Path.iterdir() as a taint-clearing enumeration the
way it does os.scandir() -- _trusted_plugin_dir's iterdir-based rebuild
still traced plugin_id through to the manifest.json open(). Switched to
scandir, matching the pattern already verified clean on PR #396.

Also stops surfacing raw exception text (update()/display() failures)
in the JSON render response -- logs full detail server-side via
exc_info instead, returning only the exception class name to the
client. And drops path values from three plugin_loader debug/error
logs that CodeQL flags as clear-text-logging of externally-influenced
data, keeping plugin_id (not flagged) for context.

* fix(dev-server): remove conditional-reassignment ambiguity in plugin_dir resolution

CodeQL's path-injection flow still traced through _parse_render_request
after the scandir fix -- the tainted find_plugin_dir() result and the
scandir-derived _trusted_plugin_dir() result shared the same variable
name (plugin_dir), reassigned only on the truthy branch. That merge
point apparently isn't treated as a barrier by the flow analysis, so it
kept tracing the pre-reassignment value through to the manifest open().

Split into two distinct names -- candidate_dir (tainted, used only to
call _trusted_plugin_dir) and trusted_dir (the only name used for any
downstream file access) -- so there's no reassigned variable for the
flow to walk through.

* fix: remove unused imports flagged by Codacy

Union in adaptive_images.py and field in adaptive_layout.py are both
imported but never used -- the last two Codacy findings on this PR,
matching the same fix already applied on PR #396.

* fix(layout): bound the fit cache; never alias the source image in fits

Two latent issues found in a self-review pass:

- LayoutContext._fit_cache was an unbounded dict (the image cache got an
  LRU cap, the text-fit cache didn't). Cache keys embed the fitted TEXT,
  so a plugin fitting changing strings — a live game clock, a ticker —
  on a 24/7 service grows it forever. Now LRU-bounded at 512 entries via
  the same pattern as the image cache.

- fit_image returned the caller's ORIGINAL image object when the source
  was already RGBA at target size (contain/fill_height, no ink crop).
  ImageFitResult is documented as an independent copy, and LayoutContext
  caches results — an aliased image lets later mutations of the source
  corrupt cached fits (or vice versa). Copy in that branch.

Both covered by new regression tests.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FqzC1nzTWL4kaqgMaQZFam

---------

Co-authored-by: Chuck <chuck@example.com>
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
 2026-07-12 10:38:52 -04:00
ChuckandClaude Sonnet 5 05e7c43b27 fix(plugins): replace dependency marker files with a real satisfaction check (#390) 
* fix(plugins): replace dependency marker files with a real satisfaction check

The .dependencies_installed hash-marker system only tracked "was this exact
requirements.txt hashed before" — not whether the packages it names are
actually present. That made it fragile (a wiped venv, a manually removed
package, or a lost/corrupted marker forces a needless full pip reinstall or,
worse, a false skip) and produced dead weight for the ~10 plugins whose
requirements.txt is comment-only (they still paid a pip subprocess on first
boot before a marker existed).

Replace it with requirements_are_satisfied() in plugin_loader.py, which
checks each real requirement line against importlib.metadata directly, so
install_dependencies() only shells out to pip when something is actually
missing or version-mismatched. Drops the marker file entirely: removed all
marker read/write sites in plugin_loader.py and store_manager.py, the
now-pointless marker-cleanup step in the git-update path, the unused legacy
marker implementation in plugin_manager.py, and the already-stale
clear_dependency_markers.sh script.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01KEZK1P1Q1fu5pcuVrkrCFZ

* fix(security): close path-injection gap in dependency-satisfaction checks

CodeQL flagged 2 new high-severity "uncontrolled data used in path
expression" alerts at the open() calls inside this PR's new
requirements_has_real_deps()/requirements_are_satisfied() -- both are
reachable from paths that were never run through the basename+trusted-base
sanitiser this codebase already uses elsewhere:

- PluginLoader.install_dependencies() only applied that sanitiser when its
  optional plugins_dir argument was actually passed; the "no plugins_dir"
  branch trusted plugin_dir_real directly. Made plugins_dir required (not
  Optional) so that branch can't exist, and added an explicit guard in
  load_plugin() so install_deps=True without a plugins_dir fails loudly
  instead of silently. Production's only real caller (PluginManager) always
  passes plugins_dir already; the harness/dev-server/render-plugin callers
  all use install_deps=False and are unaffected.

- StoreManager._install_dependencies() never sanitised plugin_path at all,
  and its call sites ultimately derive that path from a plugin's own
  manifest.json "id" field (install_plugin_from_url) -- a malicious plugin
  could otherwise point requirements_file outside plugins_dir. Applied the
  same os.path.basename()-based containment pattern PluginLoader already
  uses (and that CodeQL recognises as a real sanitiser).

Added test_install_dependencies_requires_plugins_dir and
test_install_dependencies_rejects_path_outside_plugins_dir to lock in the
actual security property, not just quiet the scanner. Verified: all 20
tests in test_plugin_loader.py pass, plus the PR's existing test plan
(test_plugin_system.py, test_store_manager_caches.py: 53 passed) and the
full CI plugin-safety suite (test_harness.py, test_visual_rendering.py,
test_plugin_matrix.py: 52 passed, 2 pre-existing skips) all still pass.

* fix(security): replace basename-only sanitiser with a trusted-enumeration check

The previous commit's os.path.basename() + os.path.join() pattern (which a
pre-existing code comment claimed CodeQL recognises as a sanitiser) did not
actually clear the alert -- the next CodeQL run still flagged the same 2
sink lines, plus a new one at the os.path.join() call itself. Taking a
substring of tainted data apparently isn't treated as a barrier by this
query, whatever the comment assumed.

Replaced it with find_trusted_subdir(): enumerate the trusted plugins_dir
via os.scandir() and only use a name that scandir itself produced, matched
by equality against the caller's requested name. The path is then built
from that enumerated entry, not from the caller's string -- a value
sourced from iterating a trusted, non-tainted directory carries no taint
regardless of what it happens to equal, which is a stronger and more
conventional allowlist-style barrier than string-stripping. Applied
identically in both PluginLoader.install_dependencies() and
StoreManager._install_dependencies(), sharing one implementation.

Re-verified: all 65 tests across test_plugin_loader.py (20, including the
2 new security regression tests), test_store_manager_caches.py (35),
test_plugin_system.py (10) pass, plus the full CI plugin-safety suite
(test_harness.py/test_visual_rendering.py/test_plugin_matrix.py: 52
passed, 2 pre-existing skips).

* fix(security): redact URL credentials from pip subprocess output before logging

CodeQL flagged 3 clear-text-logging-of-secrets alerts in
install_requirements_file() (src/common/permission_utils.py:353,360,371).
Pre-existing on main, unrelated to this PR's own diff, but now visible
since the path-injection alerts that previously took priority in the
annotation list are fixed.

The underlying risk is real: pip can echo a private index URL's embedded
basic-auth credentials (from a requirements.txt --index-url line or
PIP_INDEX_URL) back verbatim in its own stderr/stdout on failure, and this
function both logs that output directly and returns it to callers --
store_manager.py's _install_dependencies() logs result.stderr from this
same function too.

Added _redact_url_credentials(), applied immediately after each of the two
subprocess.run() calls (mutating result.stderr/stdout in place) rather
than patching each log call site individually. This closes the leak at
the source: every downstream use -- the three flagged log lines, the
"note" string embedded in the returned stdout, and store_manager.py's own
logging of the returned result -- gets the redacted text for free.

Verified the fixed-phrase "denied" check (`"a password is required" in
result.stderr`) is unaffected, since URL syntax and those phrases don't
overlap -- covered explicitly by
test_does_not_touch_denied_check_phrases. Added
test/test_permission_utils.py (6 tests) covering the redaction helper
directly and both subprocess.run() call sites (the sudo-wrapper branch,
which this repo's scripts/fix_perms/safe_pip_install.sh makes live, and
the no-wrapper fallback branch). All pass.

* fix(security): stop interpolating req_file/pip-output into log calls

The previous commit's redaction (mutating result.stderr/stdout right after
each subprocess.run()) didn't clear CodeQL's clear-text-logging alerts --
same lesson as the path-injection fix earlier in this PR: a static
analyzer can't tell "this value was already sanitised two lines up" from
"this is still the raw tainted value" just by looking at a single log
call in isolation, so it conservatively keeps flagging it regardless of
what the redaction function actually does.

Removed all dynamic interpolation (req_file, result.stderr) from the 3
flagged logger.warning() calls entirely, replacing them with fixed
messages plus (for the one that had it) result.returncode, which is a
plain int with no possible taint. The full redacted detail is still
available where it actually matters -- in the returned
CompletedProcess.stderr/stdout and the "note" text -- just not duplicated
into a log line a scanner has to reason about in isolation.

Re-verified: all 6 test_permission_utils.py tests still pass (they assert
on the returned result, not log call arguments), plus the full
test_plugin_loader.py/test_store_manager_caches.py/test_plugin_system.py
suite (71 passed, 1 pre-existing deselect, 4 subtests).

---------

Co-authored-by: Claude Sonnet 5 <noreply@anthropic.com>
 2026-07-11 08:55:50 -04:00
ChuckandClaude Sonnet 5 cbb8ec41e8 Fix: plugin/base requirements installed as web-user, invisible to root-run display service (#380) 
* fix: install plugin/base requirements as root so ledmatrix.service can see them

ledmatrix-web.service runs as a non-root user, so "Reinstall plugin
requirements" installed packages into that user's ~/.local site-packages.
ledmatrix.service (the actual display, which loads and runs plugin code)
runs as root and can't see another user's user-site packages, so plugins
with dependencies not already present system-wide would silently fail at
runtime with ModuleNotFoundError even after a "successful" reinstall.
Reproduced and fixed live against a real device (weather plugin's astral
dependency, used for moon-phase data): confirmed the exact failure
("No module named 'astral'" on every almanac cycle) and confirmed it's
gone after this fix.

Adds scripts/fix_perms/safe_pip_install.sh, a root-owned wrapper (mirroring
the existing safe_plugin_rm.sh pattern) that validates the target is
requirements.txt at the project root or under plugin-repos/ or plugins/
before running pip install as root. configure_web_sudo.sh provisions a
narrowly-scoped sudoers rule for it. api_v3.py's install_base_requirements
and install_plugin_requirements actions now use it via `sudo -n`, falling
back to today's current-user-only install (with an explanatory note) if
the wrapper isn't set up yet, so existing installs don't regress.

Also uses --ignore-installed in the wrapper: root's site-packages often has
apt/dpkg-managed copies of common libraries (requests, etc.) with no pip
RECORD file, which pip refuses to upgrade in place and aborts the *entire*
requirements.txt install over — discovered this while testing the fix live,
since a plugin's other already-satisfied-for-the-web-user dependencies had
never actually been attempted as root before.

Also fixes a pre-existing bug in configure_web_sudo.sh where the
display_controller.py/start_display.sh/stop_display.sh sudoers entries used
PROJECT_DIR (scripts/install/, where this script lives) instead of
PROJECT_ROOT (where those files actually live) — visible as the script's
own "File access test" self-check failing. Verified fixed live.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01KEZK1P1Q1fu5pcuVrkrCFZ

* fix: invoke safe_pip_install.sh via explicit bash, matching sudoers rule

CodeRabbit caught this on review: the sudoers rule configure_web_sudo.sh
provisions is scoped to "$BASH_PATH $SAFE_PIP_INSTALL_PATH *" (matching
the existing safe_plugin_rm.sh precedent in
src/common/permission_utils.py), but _pip_install_requirements() called
`sudo -n <wrapper> <req_file>` directly, relying on the script's shebang
instead of an explicit bash prefix. sudo matches the literal command line,
so this never matched the allowlisted rule on an install with only the
specific sudoers entries this script provisions — it silently fell back
to the non-root install path every time, which is the exact bug this PR
set out to fix.

This wasn't caught by live testing on ledpi.local because that device
also has a broader, non-standard "NOPASSWD: ALL" grant which masked the
mismatch. Confirmed the fix is correct by reading sudo's documented
command-matching semantics and mirroring the already-proven-working
bash-prefix pattern from permission_utils.py's safe_plugin_rm.sh call
exactly.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01KEZK1P1Q1fu5pcuVrkrCFZ

* fix: harden safe_pip_install.sh invocation against bash-path drift + address CodeRabbit nitpicks

CodeRabbit follow-up findings on the bash-prefix fix (7558aaab):

1. (Actionable) shutil.which('bash') at runtime could in principle resolve
   to a different absolute path than configure_web_sudo.sh's `command -v
   bash`, which is resolved once at setup time and baked into the static
   sudoers file as a literal string — sudo requires an exact match. Now
   tries /usr/bin/bash and /bin/bash (the standard Debian/Raspberry Pi OS
   locations, matching what the setup script virtually always produces)
   before falling back to this process's own PATH resolution, so a
   divergence in just one of them doesn't break the install.

2. (Nitpick) Any nonzero returncode was treated as "sudo denied", so a
   real pip failure (bad package, build error) would trigger a pointless
   duplicate non-root install attempt and a misleading error message.
   Now distinguishes "sudo -n rejected this exact command line" from
   "sudo ran it but the command itself failed" via sudo's own diagnostic
   text, and surfaces genuine failures immediately without retrying other
   bash candidates or falling back.

3. (Nitpick) Added structured logging for every fallback/failure path
   (wrapper missing, sudo denied, real install failure), previously only
   visible via the returned stdout note — needed for remote debugging on
   a headless Pi.

Verified: function-level smoke test confirms a real failure (this sandbox's
system python3 lacking pip) is now correctly classified as non-denial and
returned immediately without retrying candidates or double-installing.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01KEZK1P1Q1fu5pcuVrkrCFZ

---------

Co-authored-by: Claude Sonnet 5 <noreply@anthropic.com>
 2026-07-06 08:04:49 -04:00
ChuckClaude Sonnet 4.6Chuck
 		639e1c3a93 fix(web): repair news ticker custom-feeds save for JSON path (#376) 
* feat(install): surface root cause of web dependency install failures

install_dependencies_apt.py previously reported only which packages
failed, not why - the actual apt/pip error was discarded (apt) or
could scroll out of the on_error log tail (pip), leaving "Step 7:
Install web interface dependencies (line 915)" as the only visible
detail.

Capture command output for each install attempt and print a compact
DEPENDENCY INSTALLATION FAILURES summary with the last lines of error
output per package. Also run the installer with `python3 -u` for
real-time, correctly-ordered logging, and widen the on_error tail from
50 to 100 lines so the summary isn't cut off.

* fix(web): repair news ticker custom-feeds save for JSON path

The JS dotToNested() helper converts indexed form fields like
feeds.custom_feeds.0.name into a dict {'0': {name:...}} rather than a
proper array. The form-data path already had fix_array_structures() to
convert those dicts back to arrays before schema validation, but the
JSON path (used by all web-UI saves) never ran that fix, so saving any
custom feed produced a schema validation error: "Expected type array,
got object".

Add _fix_json_arrays() immediately after schema loading on the JSON
path, mirroring the existing fix_array_structures() logic.

Also fix custom-feeds.js getValue() to omit the logo key entirely when
no logo is present instead of returning logo:null, which would fail
schema validation (logo expects type object).

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix(install,tools): address PR 376 review findings

- first_time_install.sh: add _clone_rpi_rgb() wrapper so retry() cleans
  up any partial rpi-rgb-led-matrix-master dir before each clone attempt
- first_time_install.sh: use apt-get -o DPkg::Lock::Timeout=180 so apt
  handles lock contention natively instead of relying solely on flock TOCTOU check
- install_dependencies_apt.py: pass DPkg::Lock::Timeout=180 to apt-get
  install to avoid failing when unattended-upgrades holds the lock
- install_dependencies_apt.py: add type annotations to all public helpers
- api_v3.py: fix install_plugin_requirements to read plugin_manager from
  api_v3 blueprint attribute instead of the always-None module variable
- tools.html: loadGitInfo() now checks r.ok before parsing JSON and
  surfaces d.status === 'error' with the server's message in the panel

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix(tools,api): address three additional review findings

- api_v3.py install_plugin_requirements: replace hardcoded plugin-repos
  fallback with config-driven resolution (plugin_system.plugins_directory),
  matching the pattern used elsewhere in the module
- api_v3.py _fix_json_arrays: recurse into converted and existing array
  elements when items.type is object, so nested numeric-keyed dicts inside
  array items are also normalized
- tools.html toolsAction: check r.ok before r.json() and recover
  gracefully from non-JSON error bodies (HTML 500 pages), consistent
  with the existing loadGitInfo guard

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

---------

Co-authored-by: Chuck <chuck@example.com>
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
 2026-06-29 14:21:05 -04:00
Ron PierceandClaude Opus 4.8 974d7ea57a fix(install): avoid apt-package uninstall failure during web dep install (#371) 
* fix(install): avoid apt-package uninstall failure on web deps

On a fresh Pi install, requests is installed via apt (python3-requests),
which ships no pip RECORD file. When pip later installs
google-api-python-client, its dependency tree pulls a newer requests and
attempts to uninstall the apt copy, failing with "uninstall-no-record-file"
and aborting the whole install at step 7 (web interface dependencies).

Add --ignore-installed to install_via_pip so pip lays the new version down
in /usr/local (shadowing the apt copy) instead of trying to remove an
apt-managed package. This resolves the failure for any transitive
dependency pip needs to upgrade over an apt-installed package, not just
requests.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

* fix(install): also pass --ignore-installed for local rgbmatrix install

Keeps the rgbmatrix pip install consistent with install_via_pip so it
won't fail trying to uninstall an apt-managed dependency either.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
 2026-06-15 14:06:56 -04:00
ChuckandChuck 5beef0aa01 Improve first-time install error diagnostics and resilience (#369) 
* fix(install): don't let outer ERR trap mask first_time_install.sh failures

set +e alone doesn't suppress bash's ERR trap, so any non-zero exit from
first_time_install.sh inside the one-shot installer immediately triggered
the outer on_error handler with a generic "Main installation, line 370"
message — before the script could report the real exit code or point to
logs/. Suspend the trap for that block so the existing if/else handling
runs instead.

* feat(install): surface root cause of web dependency install failures

install_dependencies_apt.py previously reported only which packages
failed, not why - the actual apt/pip error was discarded (apt) or
could scroll out of the on_error log tail (pip), leaving "Step 7:
Install web interface dependencies (line 915)" as the only visible
detail.

Capture command output for each install attempt and print a compact
DEPENDENCY INSTALLATION FAILURES summary with the last lines of error
output per package. Also run the installer with `python3 -u` for
real-time, correctly-ordered logging, and widen the on_error tail from
50 to 100 lines so the summary isn't cut off.

* feat(install): harden first-time install against common Pi failure modes

- wait_for_apt_lock: apt_update/apt_install now wait (up to 3min) for
  unattended-upgrades to release the dpkg lock instead of failing
  outright with "Command failed after 3 attempts" right after first boot.
- check_disk_space: new pre-flight check (Step 1) so a full SD card fails
  fast with a clear message instead of a cryptic mid-build error.
- Step 6: wrap rpi-rgb-led-matrix git clone/submodule operations in retry
  for resilience to transient network issues.
- Step 6: capture `pip install .` build output and print the last 50
  lines on failure, so the actual cmake/compiler error is visible instead
  of just "Failed to install rpi-rgb-led-matrix Python package".

* fix(install): bound subprocess output and dedupe apt update in dependency installer

Address coderabbitai review on PR #369:
- _run() now streams combined stdout/stderr to a temp file and returns
  only the last ERROR_TAIL_LINES lines, instead of buffering full
  output in memory (Codacy also flagged the previous capture_output
  call as a subprocess-without-static-string security issue; the new
  call is annotated as safe since cmd is built from hardcoded args).
- `apt update` now runs once in main() instead of once per package
  needing an apt fallback.

* fix(install): suppress remaining Codacy subprocess false-positive

Codacy's Semgrep-based check still flagged the cmd-built subprocess.run
call as "without a static string" even with the Bandit nosec applied.
Add a nosemgrep marker alongside it - cmd is always a hardcoded
apt/pip argument list, never user input.

* fix(install): correctly detect already-installed dateutil/websocket-client

Address remaining coderabbitai findings on PR #369:
- check_package_installed() did __import__(package_name) directly, but
  python-dateutil and websocket-client import as dateutil/websocket. Both
  always failed the "already installed" check and were reinstalled on
  every run. Add an IMPORT_NAME_MAP for the mismatched names.
- _run() still read the entire temp file into memory before slicing the
  tail. Stream it line-by-line into a deque(maxlen=ERROR_TAIL_LINES)
  instead so memory use stays bounded for very chatty commands.

---------

Co-authored-by: Chuck <chuck@example.com>
 2026-06-11 18:12:35 -04:00
Ron PierceandClaude Opus 4.8 bc027c921d fix: check_plugin.py honors per-plugin test/harness.json (#365) 
check_one() always compares the render against committed golden images, but
the CLI never loaded the plugin's test/harness.json — so the deterministic
settings the goldens were generated with (config, mock data, frozen time,
sizes) weren't applied. For any time/data-dependent plugin this means the CLI
(and the plugins-repo CI workflow that calls it) renders live data and the
golden drifts on every run, even with no real regression. The pytest matrix
path already reads harness.json via load_harness_spec; the CLI now does too.

- check_one loads load_harness_spec(plugin_dir) and layers it under explicit
  CLI flags: config = schema defaults < harness.json < --config; sizes =
  --sizes > LEDMATRIX_TEST_SIZES env > harness.json > default sample;
  mock_data/freeze_time/skip_update fall back to harness.json when not given
  on the CLI.
- parse_sizes returns None (not DEFAULT_TEST_SIZES) when --sizes is omitted,
  so the env/harness.json/default fallback chain in resolve_test_sizes applies.
- Regression tests: harness.json supplies render settings, and CLI flags
  override it. Use a temp fixture plugin so they run in core CI (no plugins).

Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
 2026-06-08 12:52:33 -04:00
Ron PierceandClaude Opus 4.8 313e35a98f Add cross-size/cross-screen plugin safety harness (#361) 
* feat(testing): add cross-size/cross-screen plugin safety harness

Render every plugin across all supported matrix sizes (64x32, 128x32,
128x64, 256x32) and every declared screen, failing on crashes, content
drawn past the panel edge, or visual drift vs committed golden images.

- BoundsCheckingDisplayManager: oversized-canvas overflow detection
- harness.py: multi-size/multi-screen render engine + golden compare
- scripts/check_plugin.py: CLI (functional+bounds, --out-dir, --update-golden,
  --freeze-time); render_plugin.py refactored onto shared loading helpers
- test/plugins/test_harness.py + test_plugin_matrix.py (parametrized,
  honors per-plugin test/harness.json; skips when no plugins present)
- MockCacheManager.cache_dir so cache-dir-using plugins load headlessly
- .github/workflows/test.yml + docs/plugin-safety-harness.md

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

* fix(testing): address PR review feedback on plugin safety harness

- check_plugin: friendly error for non-numeric --sizes; reject non-object
  --config / --mock-data JSON; sanitize plugin mode before using as a
  filename; stop --update-golden from masking crash/overflow failures
- bounds_display_manager: pad the canvas out to the largest supported panel
  (not a fixed 16px) so far-overshoot coordinates are caught, not clipped
- harness: merge config_schema defaults inside render_plugin_matrix; surface
  update() failures as a non-fatal warning + result field instead of a debug
  log; sanitize mode in golden_path
- loading: fail fast when harness.json references a missing mock_data fixture
- mocks: clean up the per-instance temp cache dir via weakref.finalize
- test_plugin_matrix: add a discovery guard that fails when
  LEDMATRIX_REQUIRE_PLUGINS=1 but none found (still skips locally); type hints
- bound test deps with upper version pins for deterministic CI

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

* feat(testing): render plugins across arbitrary panel sizes, not a fixed list

Addresses maintainer feedback that there is no canonical set of supported
panel sizes — a build can be any size/configuration (square, 2x2, 4x4, 8x2,
long strips, tall stacks).

- sizes.py: SUPPORTED_SIZES -> DEFAULT_TEST_SIZES (back-compat alias kept),
  reframed as a representative SAMPLE of real panel-grid arrangements rather
  than an authoritative list; add parse_size_token / coerce_sizes /
  resolve_test_sizes helpers
- sizes are now fully overridable: LEDMATRIX_TEST_SIZES env (global, e.g. test
  on your exact hardware) > per-plugin harness.json "sizes" > default sample;
  CLI --sizes unchanged
- bounds_display_manager: pad the canvas to the largest panel IN THE CURRENT
  RUN (via overflow_extent) instead of a hardcoded max, so cross-size overflow
  detection scales to whatever sizes a run uses
- harness: compute per-run extent and thread it into the bounds manager
- tests: arbitrary-shape + size-parsing/precedence coverage
- docs: rewrite "Supported sizes" -> "Sizes: a sample, not a fixed list"

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

* fix(testing): fail the harness on non-connectivity update() errors

Addresses the remaining review thread: recording every update() exception as a
non-fatal warning still let a real update() regression pass green as long as
display() survived. Now update() failures are classified — a tolerated set of
connectivity errors (ConnectionError/TimeoutError/socket/ssl/urllib/http/
requests) is recorded non-fatally (expected with no network in CI), while any
other exception is treated as a genuine bug and fails that render.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

* ci(security): pin actions to SHAs and disable checkout credential persistence

Addresses the CodeRabbit/zizmor workflow-hardening finding: pin
actions/checkout and actions/setup-python to full commit SHAs and set
persist-credentials: false on checkout to reduce supply-chain and
token-exposure risk.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

* fix(testing): validate positive sizes; narrow requests import except

Two review findings:
- sizes.py: parse_size_token / coerce_sizes now reject non-positive
  dimensions (0x32, -64x32) with a clear message instead of passing invalid
  sizes downstream (CodeRabbit).
- harness.py: the optional `requests` import now catches ImportError
  specifically and logs instead of `except Exception: pass`, clearing the
  Codacy medium "Try, Except, Pass" (harness.py L52) and Ruff S110/BLE001.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
 2026-06-05 14:32:52 -04:00
ChuckClaude Sonnet 4.6Chuck
 		713539e491 fix(web-ui): fix quick actions not firing, add toast feedback, suppress install handler warning (#346) 
* fix(web-ui): fix quick actions not firing, add toast feedback, suppress install handler warning

- base.html: add htmx:afterSettle listener to set data-loaded on tab
  containers after HTMX swaps their content, preventing the overview
  partial from being re-fetched (and handlers lost) on every tab switch
- base.html: call htmx.process() in loadOverviewDirect/loadPluginsDirect
  fallbacks so buttons get HTMX handlers even if HTMX finished its
  initial body scan before the fallback fetch completed
- overview.html + index.html (11 buttons): replace event.detail.xhr.responseJSON
  (undefined in HTMX 1.9.x) with JSON.parse(event.detail.xhr.responseText)
  so quick action toast notifications actually fire
- plugins_manager.js: add guarded htmx:afterSettle listener that only calls
  attachInstallButtonHandler when #install-plugin-from-url is in the DOM,
  eliminating the spurious console warning on non-plugin tab loads

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix(web-ui): ensure quick-action toasts always fire even on xhr/parse failure

Replace silent catch(e){} in all 11 hx-on:htmx:after-request handlers with a
pattern that sets default message/status before the try block and calls
showNotification(m,s) unconditionally after it, so a fallback toast is shown
whenever xhr is absent or responseText is not valid JSON.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix(web-ui): show error toast on non-JSON 4xx/5xx quick-action responses

In the catch block of all 11 hx-on:htmx:after-request handlers, check
xhr.status >= 400 and downgrade s to 'error' so a failed action that
returns an HTML error page (or other non-JSON body) surfaces as an error
toast instead of the optimistic 'success'/'info' default.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix(web-ui): guard setTimeout fallback for attachInstallButtonHandler

The 500ms fallback setTimeout was calling attachInstallButtonHandler()
unconditionally even when the plugins partial wasn't in the DOM, causing
a spurious console.warn on every page load. Add the same element-existence
check already present on the htmx:afterSettle listener.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* Fix backup API 404s, hardware status 500, and HTMX loading race

- Add all backup API routes to api_v3.py: preview, list, export,
  validate, restore (with plugin reinstall), download, delete
- Fix PermissionError on /hardware/status: return graceful 200 instead
  of 500 when the status file is owned by a different user; also fix
  root cause by writing the file world-readable (0o644) in display_manager
- Fix HTMX race: dispatch htmx:ready window event from HTMX onload
  callback; loadTabContent now waits for that event instead of
  immediately falling back to direct fetch (eliminating the
  "HTMX not available" console warning on initial load)

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* Cancel HTMX fallback timers when htmx:ready fires

The 5-second setTimeout fallbacks for plugins and overview were firing
before the htmx:ready event arrived, logging spurious warnings. Each
timer now self-cancels via htmx:ready so the fallback only triggers
when HTMX genuinely fails to load.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* Address review feedback: error leaks, ok:false, htmx:ready coverage

- Backup endpoints: replace raw str(e) in user-facing responses with a
  generic message; full exception still logged via exc_info=True
- hardware/status: change ok:null to ok:false for PermissionError and
  json.JSONDecodeError so the UI's hw.ok===false check triggers correctly
- base.html: dispatch htmx:ready from the fallback load path so any
  deferred listeners fire on CDN-fallback loads too
- loadTabContent: also listen for htmx-load-failed so overview/wifi/plugins
  fall back to direct fetch when HTMX is completely unavailable

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* Treat system-managed pip packages as satisfied for dependency marker

When a plugin's requirements.txt includes a package installed via the
system package manager (dnf/apt), pip fails with 'uninstall-no-record-file'
because it can't replace the system-tracked copy. The package is present
and functional, but the missing marker caused the install to be retried
on every service restart.

Detect this specific error pattern: if the only pip failure is
uninstall-no-record-file, write the .dependencies_installed marker and
log a warning instead of returning False, suppressing the repeated warning.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* Fix uninstall-no-record-file detection condition

The previous check used a string replacement that left 'error:' in the
remaining text, causing the condition to always evaluate false. Simplify
to a direct substring check: if 'uninstall-no-record-file' appears in pip
stderr the affected package is installed at the system level and we write
the marker, suppressing the repeated warning on every restart.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* Resolve CodeQL security findings in backup API

Path traversal (CWE-22):
- backup_download: switch from send_file(user-tainted-path) to
  send_from_directory(_BACKUP_EXPORT_DIR, filename); Flask uses
  werkzeug safe_join internally which CodeQL recognises as a sanitizer
- backup_delete: enumerate the export directory and match by name so
  entry.unlink() operates on a filesystem-derived Path rather than one
  constructed from user input; _safe_backup_path still guards first

Information exposure through exceptions (CWE-209):
- backup_validate: err_msg from validate_backup() can embed exception
  strings containing temp-file paths; log the detail, return a generic
  'Invalid or corrupted backup file' to the client
- Other backup endpoints: already fixed (str(e) -> generic message);
  CodeQL alerts will clear on next scan

plugin_loader.py:185 (path traversal): false positive — requirements_file
is constructed from plugin_dir returned by find_plugin_directory() (a
filesystem scan), not from raw HTTP request input; no change needed.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* Fix pre-existing information exposure in version and action endpoints

- get_system_version (alert #218): replaced str(e) with generic message;
  exception still logged via logger.error(exc_info=True)
- execute_system_action (alert #216): removed str(e) and full
  traceback.format_exc() from the HTTP response — the full stack trace
  was being sent directly to clients; replaced with generic message and
  proper logger.error call

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* Fix remaining GitHub CodeQL security alerts

- py/stack-trace-exposure: Remove str(e) and traceback.format_exc() from
  all HTTP responses across api_v3.py, pages_v3.py, and app.py; replace
  with generic messages and logger.error(exc_info=True)
- py/reflective-xss: Escape partial_name via markupsafe.escape in the
  load_partial 404 response
- py/path-injection: Add regex validation of plugin_id before filesystem
  use in _load_plugin_config_partial
- py/incomplete-url-substring-sanitization: Replace 'github.com' in
  substring checks with urlparse hostname comparison in store_manager.py
- py/clear-text-logging-sensitive-data: Remove football-scoreboard debug
  prints and sensitive request-body prints from update endpoint
- js/bad-tag-filter: Replace script-only regex in BaseWidget.sanitizeValue
  with DOM-based textContent stripping that removes all HTML
- js/incomplete-sanitization: Fix escapeAttr to properly encode &, ", ',
  <, > using HTML entities instead of backslash escaping
- js/prototype-pollution-utility: Add __proto__/constructor/prototype
  key guards to deepMerge function in plugins_manager.js
- app.py error handlers: Always return generic messages; remove debug-mode
  branches that could expose tracebacks in production

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* Fix three remaining CodeQL path-injection and info-exposure alerts

- plugin_loader.py: resolve plugin_dir with strict=True and validate
  marker_path with relative_to() before any filesystem writes, giving
  CodeQL the positive sanitization pattern it requires (py/path-injection)
- api_v3.py _safe_backup_path: replace substring negative checks with a
  strict positive regex (^[a-zA-Z0-9][a-zA-Z0-9._-]{0,200}\.zip$) that
  CodeQL recognises as sanitising the user-supplied filename
  (py/path-injection)
- api_v3.py backup_validate: whitelist known-safe manifest fields before
  returning JSON, preventing any exception strings captured inside
  validate_backup() from reaching the HTTP response (py/stack-trace-exposure)

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* Resolve 29 open CodeQL security alerts across 5 files

py/flask-debug (#214):
- debug_web_manual.py: read debug mode from LEDMATRIX_FLASK_DEBUG env var
  instead of hardcoded True

py/stack-trace-exposure (#216, #218):
- api_v3.py execute_system_action: remove subprocess stdout/stderr from
  HTTP responses; log via logger instead
- api_v3.py get_git_version: validate output matches safe ref format
  (^[a-zA-Z0-9._-]+$) before including in response
- api_v3.py: remove all remaining traceback.format_exc() dead variables
  and print() debug calls (replaced with logger.debug/warning)

py/reflective-xss (#207, #208, #209, #210, #211, #212):
- api_v3.py: remove plugin_id from all error/success response messages
  (uninstall, install, update, health, not-found responses)
- pages_v3.py load_partial: return static "Partial not found" message
  instead of echoing partial_name
- pages_v3.py _load_starlark_config_partial: add app_id regex validation,
  use static error messages instead of f-strings with app_id

py/path-injection (#187–#206):
- pages_v3.py _load_plugin_config_partial: resolve plugins_base and
  validate _plugin_dir with relative_to() before all file operations;
  same for assets metadata directory
- pages_v3.py _load_starlark_config_partial: resolve starlark_base and
  validate schema_file/config_file paths with relative_to()
- plugin_loader.py _find_plugin_directory: resolve plugins_dir and
  validate strategy-2 candidates with relative_to()
- plugin_loader.py install_dependencies: resolve plugin_dir first, then
  construct requirements_file and marker_path from resolved base
- plugin_loader.py load_module: resolve plugin_dir with strict=True and
  validate entry_file with relative_to() before exec_module

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* Fix 15 remaining CodeQL path-injection and stack-trace-exposure alerts

Switch from resolve()+relative_to() to os.path.basename() reassignment,
which CodeQL recognizes as a path sanitizer that breaks the taint chain.
Also remove exception objects from backup_manager validate_backup return
strings to eliminate the stack-trace-exposure taint source.

Fixes alerts #227, #233, #234, #235, #237, #238, #239, #240, #241,
#242, #243, #244, #245, #246, #247.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* Fix broken logger format string and leaked exception in config save error

- pages_v3.py: plain string was used instead of %-style substitution,
  so every manifest-read failure logged the literal "{plugin_id}"
- api_v3.py save_main_config: exception message was still leaking
  through the error response; replace with generic message (consistent
  with the rest of the CodeQL sweep in this PR)

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

---------

Co-authored-by: Chuck <chuck@example.com>
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
 2026-05-24 09:29:53 -04:00
ChuckChuckClaude Sonnet 4.6
 		d941c91f24 fix(systemd): wait for network connectivity before starting services (#335) 
Change After=network.target → After=network-online.target + Wants=network-
online.target in both service templates and install_web_service.sh.

network.target only guarantees NetworkManager has started — it does NOT
mean the device has an active internet connection. On boot the LED matrix
service was starting within seconds of the network interface appearing,
before WiFi association and DHCP completed, causing every first-update API
call to fail with "Network is unreachable" or DNS resolution errors.

network-online.target waits for a confirmed route before the service fires.
On Raspberry Pi OS this is provided by NetworkManager-wait-online. The
tradeoff is a few extra seconds at boot, acceptable for a display device.

Observed on devpi: service started at 14:48:03, all API calls (weather,
FlightRadar24, local ADS-B) failed at 14:48:07 with network errors, then
the service restarted cleanly at 14:50:40 once WiFi was established.

Co-authored-by: Chuck <chuck@example.com>
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
 2026-05-15 15:47:35 -04:00
ChuckClaude Sonnet 4.6Chuck
 		05b3fa56cb fix: Codacy security fixes, CVE dependency bumps, and code quality cleanup (#331) 
* fix(deps): bump minimum versions to address CVEs

Pillow 10.4.0 → 12.2.0: CVE-2026-40192 (DoS via FITS decompression bomb),
CVE-2026-25990 (OOB write via PSD image), CVE-2026-42311/42308/42310

requests 2.32.0 → 2.33.0: CVE-2026-25645 (temp file security bypass),
CVE-2024-47081 (.netrc credentials leak)

werkzeug 3.0.0 → 3.1.6: CVE-2023-46136, CVE-2024-49766/49767,
CVE-2025-66221, CVE-2026-21860/27199 (DoS, path traversal, safe_join bypass)

Flask 3.0.0 → 3.1.3: CVE-2026-27205 (session data caching info disclosure)

spotipy 2.24.0 → 2.25.2: CVE-2025-27154, CVE-2025-66040

python-socketio 5.11.0 → 5.14.0: CVE-2025-61765

pytest 7.4.0 → 9.0.3: CVE-2025-71176 (insecure temp dir handling)

Updated in requirements.txt, web_interface/requirements.txt,
plugin-repos/starlark-apps/requirements.txt, and
plugin-repos/march-madness/requirements.txt.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix: resolve Pylint errors in executor, data service, and odds call

Rename TimeoutError to PluginTimeoutError in plugin_executor.py to
avoid shadowing the built-in; no external callers affected.

Remove dead try/except in BackgroundDataService.shutdown: executor.shutdown()
never accepted a timeout kwarg so the try branch always raised TypeError.
Simplify to a direct shutdown(wait=wait) call.

Remove is_live kwarg from odds_manager.get_odds() call in sports.py;
BaseOddsManager.get_odds() has no such parameter. The live update interval
is already encoded in the update_interval_seconds argument passed alongside.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix: MD5→SHA-256, shellcheck warnings, and broken doc links

config_service.py: replace MD5 with SHA-256 for config change detection;
same semantics (equality comparison), no stored hashes affected.

Shell scripts — shellcheck warnings:
- diagnose_web_interface.sh: remove useless cat (SC2002)
- dev_plugin_setup.sh: restructure A&&B||C into if/then (SC2015)
- fix_assets_permissions.sh: remove unused REAL_HOME block (SC2034)
- install_web_service.sh: remove unused USER_HOME assignment (SC2034)
- diagnose_web_ui.sh: remove unused SUDO assignments (SC2034)
- diagnose_plugin_permissions.sh: remove unused BLUE color var (SC2034)
- first_time_install.sh: remove unused CLEAR var, PACKAGE_NAME
  assignment, and replace loop variable with _ (SC2034)

docs/PLUGIN_ARCHITECTURE_SPEC.md: fix 10 broken TOC anchor links to
include section numbers matching the actual headings (MD051).

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix: remove unused imports and bare exception aliases (pyflakes F401/F841)

Remove unused imports across 86 files in src/, web_interface/, test/,
and scripts/ using autoflake. No logic changes — only dead import
statements and unused names in from-imports are removed.

Also remove bare exception aliases where the variable is never
referenced in the handler body:
- src/cache/disk_cache.py: except (IOError, OSError, PermissionError) as e
- src/cache_manager.py: except (OSError, IOError, PermissionError) as perm_error
- src/plugin_system/resource_monitor.py: except Exception as e
- web_interface/app.py: except Exception as read_err

86 files changed, 205 lines removed, 18 pre-existing test failures unchanged.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix: remove unused local variable assignments (pyflakes F841)

Dead assignments removed across src/ and web_interface/:

- background_data_service: drop future= on fire-and-forget executor.submit
- base_classes/baseball: drop font= (all rendering uses self.fonts['time'])
- base_classes/hockey: drop status_short= (never referenced after assignment)
- common/cli: drop game_helper=/config_helper= bindings in import-test block;
  constructors called for instantiation-only validation
- common/display_helper: drop text_width= (x_position uses display_width
  directly); drop draw= in create_error_image (uses _draw_centered_text)
- config_manager: remove dead secrets_content loading block in migration path
  (comment already noted save_config_atomic handles secrets internally)
- display_manager: drop setup_start= (timing was never completed or read)
- font_manager: drop target_path= (catalog uses font_file_path directly);
  drop face=/font= bindings in validate_font (validation by construction —
  TypeError on failure is the signal, not the return value)
- font_test_manager: drop width=/height= (draw_text uses display_manager directly)
- plugin_system/state_reconciliation: drop manager= (only config/disk/state_mgr used)
- plugin_system/store_manager: drop result= on pip install subprocess.run
  (check=True raises on failure; stdout unused)
- web_interface/blueprints/pages_v3: drop main_config_path=""/secrets_config_path=""
  (render_template uses config_manager.get_*_path() inline)

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix(js): resolve ESLint no-undef warnings across 6 JS files

Three distinct patterns:

1. Vendor library globals — htmx is injected by <script> before these
   extension files load; ESLint lints files in isolation and doesn't know.
   Fix: add /* global htmx */ to htmx-sse.js and htmx-json-enc.js.

2. Cross-file globals — showNotification is defined as window.showNotification
   in app.js/notification.js but called bare in app.js and error_handler.js.
   ESLint doesn't connect window.X = Y with a bare call to X.
   Fix: add /* global showNotification */ to app.js and error_handler.js.

3. Forward-reference window.* functions — in array-table.js, checkbox-group.js,
   and custom-feeds.js, functions like removeArrayTableRow are called early
   inside event-handler closures but assigned to window.* later in the file.
   At runtime this works (the handler fires after the assignment), but ESLint
   sees the bare name at the call site.
   Fix: change bare calls to window.removeArrayTableRow(this) etc. so the
   reference is explicit and ESLint-safe.

Also guard the updateSystemStats call in app.js reconnectSSE: the function
is called but defined nowhere in the codebase. Guard with typeof check so
it won't throw ReferenceError if the reconnect path is hit.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix(js): resolve Biome lint warnings across 9 JS files

noUnusedVariables (catch bindings → optional catch syntax):
- app.js, file-upload.js, timezone-selector.js: } catch (e) { → } catch {
  ES2019 optional catch binding; e was unused in all three handlers

noUnusedVariables (dead assignments):
- app.js: remove const data= in display SSE stub (handler does nothing yet)
- api_client.js: remove const timeoutId= (setTimeout ID never used to cancel)
- custom-feeds.js: remove const oldIndex= (getAttribute result never read)
- schedule-picker.js: remove const compactMode= (never used in HTML build)
- select-dropdown.js: remove const icons= (icons not yet rendered in options)

noPrototypeBuiltins:
- day-selector.js: DAY_LABELS.hasOwnProperty(x) →
  Object.prototype.hasOwnProperty.call(DAY_LABELS, x)
  Safe form that works even on null-prototype objects

useIterableCallbackReturn:
- file-upload.js, notification.js: forEach(x => expr) →
  forEach(x => { expr; }) — forEach ignores return values;
  implicit return from arrow body was misleading

htmx-sse.js is a vendor extension file with old-style var/== patterns
that are correct for it; 18 Biome issues suppressed via Codacy API
rather than modifying the vendor source.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix(security): escape user input in raw HTML responses in pages_v3.py

plugin_id comes directly from the URL path
(/partials/plugin-config/<plugin_id>) and was interpolated into an HTML
fragment without escaping. A crafted URL like
/partials/plugin-config/<script>alert(1)</script> would inject that
tag into the DOM via the HTMX partial response.

Fix: wrap all user-controlled values in markupsafe.escape() before
embedding in raw HTML strings. Affects the plugin-not-found 404
response and both error 500 responses in the plugin config partial.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix: address Bandit B108/B110 across production code

B110 (try/except/pass):
- display_controller.py: narrow 'except Exception' to 'except AttributeError'
  for get_offset_frame() — plugins not having this optional method is the
  expected case, not all exceptions
- config_manager.py: B110 already resolved by the earlier removal of the
  dead secrets-loading block (the except/pass was inside it)
- All other except/pass blocks in src/ and web_interface/ are intentional
  (last-resort recovery, best-effort fallbacks, non-critical startup probes).
  Annotated each with # nosec B110 and a brief inline reason so the decision
  is explicit for future reviewers.
- Test files and plugin-repos B110 suppressed via Codacy API (not prod code).

B108 (/tmp usage):
- permission_utils.py: /tmp listed to PREVENT permission changes on it — not
  used as a temp path. Annotated # nosec B108.
- display_manager.py: fixed snapshot path is intentional (web UI reads same
  path); path-check guard also annotated.
- wifi_manager.py: named /tmp files match the sudoers allowlist installed with
  the system (the paths are hard-coded in both places by design). Annotated
  all six open/cp references # nosec B108.
- scripts/render_plugin.py: dev script default overridable by user. Annotated.
- web_interface/app.py: reads the same fixed path written by display_manager.
  Annotated # nosec B108.
- Test files suppressed via Codacy API.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix: address remaining Codacy security findings

Flask debug=True (real fix):
- web_interface/app.py: debug=True in __main__ block exposes the Werkzeug
  interactive debugger (arbitrary code execution). Changed to
  os.environ.get('FLASK_DEBUG', '0') == '1' — off by default, opt-in
  via environment variable for local development.

nosec annotations (accepted risk with documented rationale):
- disk_cache.py: os.chmod(0o660) is intentional — web UI and LED matrix
  service share a group, 660 gives group write while denying world access
  (B103 + Semgrep insecure-file-permissions suppressed in Codacy)
- wifi_manager.py: urlopen to hardcoded connectivity-check.ubuntu.com URL
  (B310 — no user input involved)
- font_manager.py: urlretrieve URL comes from user's own config file on
  their local device (B310)
- start_web_conditionally.py: os.execvp with both sys.executable and a
  fixed PROJECT_DIR-relative constant (B606)

Confirmed false positives suppressed via Codacy API (15 issues):
- SSRF (3x): client-side JS fetch — SSRF is server-side; browser fetch
  is CORS-restricted to same origin
- B105 (3x): test fixtures use dummy secrets by design; store_manager
  checks for the placeholder string, it is not itself a secret
- PMD numeric literal (2x): 10000000 is within Number.MAX_SAFE_INTEGER
- Prototype pollution (1x): read-only schema traversal, no writes
- no-unsanitized_method (1x): dynamic import() is CORS-restricted
- detect-unsafe-regex (1x): operates on server-controlled config values
- plugin-repos B103 (1x): vendor code chmod on executable
- Semgrep insecure-file-permissions (3x): same disk_cache 0o660 as above

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix: remove unnecessary f prefix from f-strings without placeholders (F541)

Pyflakes F541 flags f-strings that contain no {} interpolation — they are
identical to plain strings but trigger unnecessary string formatting overhead.

Fixed in production code:
- src/base_classes/data_sources.py (2 debug log calls)
- src/logo_downloader.py (1 error log)
- src/plugin_system/store_manager.py (5 strings across 3 log calls)
- src/web_interface/validators.py (1 return value)
- src/wifi_manager.py (4 log/message strings)
- web_interface/start.py (1 print)

F541 issues in test/, scripts/, and plugin-repos/ suppressed via Codacy API
as non-production code.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* chore(dev): add Pillow compatibility smoke test script

Covers all Pillow APIs used in LEDMatrix — image creation, drawing,
font metrics, LANCZOS resampling, paste/alpha_composite, and PNG I/O.
Run after any Pillow version bump to catch regressions before deploy.

    python3 scripts/dev/test_pillow_compat.py

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix: resolve 8 new Codacy issues introduced by PR changes

shellcheck SC2034:
- first_time_install.sh: 'type' loop variable also unused in the wifi
  status loop (we previously fixed 'device' → '_' but left 'type').
  Changed to '_ _ state' since neither device nor type is referenced.

ESLint no-undef:
- app.js: typeof guards don't satisfy no-undef; added updateSystemStats
  to the /* global */ declaration alongside showNotification.

nosec annotation:
- web_interface/app.py: app.run(host='0.0.0.0') line changed when we
  fixed debug=True, giving it a new issue ID. Re-added # nosec B104.

pyflakes F401:
- scripts/dev/test_pillow_compat.py: ImageFilter was imported but never
  used in the smoke test. Removed from the import.

Codacy API suppressions (false positives on changed lines):
- disk_cache.py 0o660 chmod (2x): lines changed when # nosec B103 was
  added, producing new Semgrep issue IDs. Re-suppressed.
- pages_v3.py raw-html-concat: Semgrep does not recognise escape() as
  a sanitizer; the escape() call IS the correct fix.
- app.py flask 0.0.0.0: same line as B104 above; Semgrep rule also
  re-suppressed.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix: address PR review findings

Fix (10 of 15 findings):

plugin-repos/march-madness/requirements.txt:
  Add urllib3>=1.26.0 — manager.py directly imports from urllib3; it was
  an undeclared transitive dependency via requests.

scripts/dev/dev_plugin_setup.sh:
  Restore subshell form (cd "$target_dir" && git pull --rebase) || true
  so the shell's working directory is not permanently changed after the
  if-cd block. Previous fix for SC2015 leaked cwd into the remainder of
  the script.

src/base_classes/sports.py:
  Narrow 'except Exception' to 'except RuntimeError as e' and log via
  self.logger.debug — Path.home() raises only RuntimeError for service
  users; other exceptions should not be silently swallowed.

src/config_service.py:
  Fix stale "MD5 checksum" in ConfigVersion.__init__ docstring (line 40);
  the implementation uses SHA-256 since the Codacy fix.

src/wifi_manager.py:
  Log the last-resort AP enable failure with exc_info=True instead of
  silently passing — failure here means the device may be unreachable.

web_interface/blueprints/pages_v3.py:
  Log the outer metadata pre-load exception at debug level instead of
  swallowing it silently; schema still loads fully below.

src/background_data_service.py:
  Remove unused 'timeout' parameter from shutdown() — executor.shutdown()
  does not accept timeout; update __del__ caller accordingly.

src/font_manager.py:
  Validate URL scheme before urlretrieve — reject non-http/https schemes
  (e.g. file://) to prevent reading local files from config-supplied URLs.

src/plugin_system/plugin_executor.py:
  Simplify redundant except tuple: (PluginTimeoutError, PluginError,
  Exception) → Exception, which already covers the others.

test/test_display_controller.py:
  Mark empty test_plugin_discovery_and_loading as @pytest.mark.skip with
  reason. Move duplicate 'from datetime import datetime' to module header
  and remove the stray mid-module copy.

Skip (5 of 15 findings, with reasons):
  - pytest 9.0.3 concerns: full suite already verified (467 pass, 18 pre-existing)
  - Pillow 12.2.0 API concerns: no deprecated APIs in codebase; tests + Pi smoke test pass
  - diagnose_web_ui.sh sudo validation: set -e already ensures fail-fast on any sudo failure
  - app.py request-logging except: must stay silent (recursive logging risk); annotated
  - app.py SSE file-read except: genuinely transient I/O; annotated

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

---------

Co-authored-by: Chuck <chuck@example.com>
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
 2026-05-15 10:19:55 -04:00
ChuckClaude Sonnet 4.6Chuck
 		3b45a75f75 fix: pixlet install false-failure, force render in web service, web UI perf (#328) 
* fix: pixlet install false-failure, force render in web service, web UI perf

Fixes three user-reported issues:

1. Pixlet install reported failure even on success — `set -e` in
   download_pixlet.sh caused the script to exit non-zero when
   `((success_count++))` evaluated the post-increment old value (0),
   which bash treats as a failed command. Fixed with shell arithmetic
   assignment instead.

2. Force render returned 503 "plugin not loaded in web service" — the
   web service runs its own PluginManager with display_manager=None,
   so the starlark-apps plugin can never be loaded there. Added a
   standalone render path (_standalone_render_starlark_app) that calls
   pixlet directly from the web service, writing to cached_render.webp.
   The main-service plugin path is preserved and tried first.

3. Web UI sluggishness — the SSE /stream/stats generator was blocking
   a Flask thread for 1s per tick via psutil.cpu_percent(interval=1).
   Switched to non-blocking interval=None (primed at startup). Also
   cached the per-client ledmatrix systemctl check (15s TTL), raised
   the AP mode check TTL from 5s to 30s, and halved the display
   preview poll rate from 0.5s to 1s.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* refactor(starlark): surface config errors, deduplicate pixlet lookup, uniform response shape

- _standalone_render_starlark_app: split silent except into separate
  json.JSONDecodeError and OSError handlers that return (False, message)
  with the file path and parse error, so callers know when config.json
  is unreadable rather than silently rendering with empty config

- get_starlark_status: replace 14-line inline platform/shutil pixlet
  lookup with _find_pixlet_binary(), which also checks the user-
  configured starlark-apps.pixlet_path — the old code never consulted
  that setting so the status endpoint could wrongly report pixlet
  unavailable even when a custom path was configured

- render_starlark_app standalone branch: add frame_count: 0 to success
  and error responses so both branches return the same payload shape;
  0 is accurate since standalone render writes cached_render.webp but
  does not extract frames into memory

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix(starlark): safe chmod fallback in pixlet lookup, semantic HTTP codes in standalone render

_find_pixlet_binary: bundled.chmod(0o755) could raise OSError (e.g.
file owned by root) and abort the entire resolution chain, skipping
the PATH fallback. Now checks os.access first; if chmod is needed,
wraps it in try/except OSError, logs a warning, and falls through to
shutil.which so PATH is always tried.

_standalone_render_starlark_app: expanded return type from
(bool, str) to (bool, int, str) so each failure mode carries a
semantic HTTP status code:
  app/star file missing  → 404
  invalid/unreadable config.json → 400
  pixlet binary missing  → 503
  pixlet non-zero exit   → 502
  subprocess timeout     → 504
  unexpected exception   → 500
  success                → 200
Call site updated to unpack (success, status_code, error) and forward
status_code directly in both success and error responses.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix(starlark): validate config.json is a dict before iterating items

json.load succeeds on valid JSON that isn't an object (e.g. an array
or bare string), leaving app_config as a non-dict and causing an
AttributeError on the subsequent .items() call. Added isinstance check
immediately after json.load; non-dict values return (False, 400, ...)
with the actual type name in the message.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix(starlark): harden pixlet binary check and manifest shape validation

_find_pixlet_binary: switched bundled.exists() to bundled.is_file() so
a directory named pixlet-linux-arm64 (traversable, hence X_OK) is no
longer returned as a binary; re-check os.access after chmod succeeds
so we only return the path when executability is confirmed, with a
separate warning if it still isn't executable after chmod.

_standalone_render_starlark_app: added isinstance guards on the
manifest and apps values returned by _read_starlark_manifest(); a
manifest.json containing valid non-object JSON (e.g. an array) would
previously raise AttributeError on .get(); invalid shape now returns
(False, 400, ...) instead.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

---------

Co-authored-by: Chuck <chuck@example.com>
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
 2026-05-13 14:14:48 -04:00
ChuckClaude Sonnet 4.6Chuck
 		1a0f1c8015 fix: service control buttons and AP-mode SSH lockout post-install (#326) 
* fix: service control buttons and AP-mode SSH lockout post-install

Two user-reported issues after fresh install:

1. All service buttons (Start/Stop/Restart Display, Restart Web Service)
   failed silently — only Reboot worked.

   Root cause: sudoers rules use `ledmatrix.service` (with suffix) but
   api_v3.py called `sudo systemctl start ledmatrix` (no suffix). sudo
   does exact string matching, so every service action was rejected with
   returncode=1. Also missing from sudoers: ledmatrix-web, journalctl,
   and is-active entries.

   Fix:
   - Add `.service` suffix to all 8 sudo systemctl call sites in
     api_v3.py (_ensure_display_service_running, _stop_display_service,
     and all execute_system_action branches).
   - Add timeout=15 to all subprocess.run calls in execute_system_action
     (previously could hang indefinitely).
   - Add missing sudoers rules to first_time_install.sh and
     configure_web_sudo.sh: ledmatrix-web.service start/stop/restart,
     is-active for both name forms, and journalctl -u/-t ledmatrix rules.

2. SSH and web UI became inaccessible after ~1 hour even though the
   display kept running.

   Root cause: wifi_monitor_daemon restarts NetworkManager after 5
   consecutive internet failures (~2.5 min). Each NM restart drops WiFi
   briefly. During that window check_and_manage_ap_mode() increments
   _disconnected_checks but the daemon never reset it after the restart.
   After 3 such NM-restart cycles, _disconnected_checks reached 3 and
   AP mode activated — changing the Pi from WiFi client to hotspot
   (192.168.4.1) and killing SSH on the old IP.

   Fix:
   - Reset wifi_manager._disconnected_checks = 0 in the daemon
     immediately after a successful NM restart so the brief drop it
     causes doesn't count toward AP-mode activation.
   - Increase _disconnected_checks_required from 3 to 6 (90s → 3min)
     as an additional buffer against transient network flaps.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* revert: restore AP-mode grace period to 90s (3 checks)

The counter reset after NM restart already fully prevents the SSH-lockout
cascade: _disconnected_checks can never accumulate across NM restarts
because it is reset to 0 before the next daemon iteration runs.

The 3→6 increase provided no additional fix for the described problem and
caused a UX regression: fresh Pi devices with no WiFi configured would
wait 3 minutes instead of 90 seconds for the LEDMatrix-Setup hotspot to
appear.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix: address five valid review findings; skip two

Fixed:
- march-madness/requirements.txt: Pillow>=10.3.0 (patches CVE-2024-28219;
  10.3.0 is the actual fix version — reviewer cited 12.2.0 but that risks
  breaking API changes without test coverage)
- wifi_monitor_daemon.py: add missing `import subprocess`; subprocess.run
  and CalledProcessError would NameError at runtime on the NM restart path
- wifi_manager.py: validate ap_idle_timeout_minutes before arithmetic —
  coerce to int, clamp 1–1440, fall back to 15 on bad config values
- wifi_manager.py: call _remove_nm_dnsmasq_captive_conf() on all three
  rollback paths in _enable_ap_mode_nmcli_hotspot() and in the top-level
  except block so stale dnsmasq drop-ins are never left behind
- api_v3.py: fix wrong_password prefix strip — removeprefix("wrong_password:")
  then lstrip() handles both "wrong_password: msg" and "wrong_password:msg"
- plugins_manager.js: add .catch() to loadInstalledPlugins().then() to
  surface failures instead of silently dropping unhandled rejections

Skipped:
- WiFiManager AP state persistence: architectural overhaul; _is_ap_mode_active()
  already derives from live system state, not in-memory variables
- Absolute subprocess paths in api_v3.py: paths vary by distro (/usr/bin vs
  /bin); web service has a normal PATH; sudoers already use resolved paths

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix: address five review findings (NM retry loop, start_display message, code quality)

- wifi_monitor_daemon: reset _consecutive_internet_failures = 0 in both
  NM-restart exception handlers; previously both left the counter at threshold,
  causing an immediate retry on the next iteration instead of waiting another
  full backoff period

- api_v3: fix start_display failure message — when mode is set and systemctl
  returns non-zero, message now includes the failure reason and a hint rather
  than always reporting success phrasing

- wifi_manager: move _redirect_backend from class variable to instance variable
  in __init__ alongside _ap_enabled_at; class-level default shadowed correctly
  in practice (single instance) but was misleading

- wifi_manager: narrow broad except Exception in _check_internet_connectivity
  to (subprocess.SubprocessError, OSError) for ping and OSError for HTTP
  (urllib.error.URLError is an OSError subclass in Python 3)

- wifi_manager: remove redundant local 'import re as _re' in _validate_ap_config;
  re is already imported at module level (line 37)

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix: address five review findings (Pillow CVEs, daemon exception narrowing, timeout handling, plugin store)

- march-madness/requirements.txt: Pillow>=12.2.0 (patches CVE-2026-42308
  and CVE-2026-42310; previous floor of 10.3.0 was insufficient)

- wifi_monitor_daemon: narrow final except Exception to
  (subprocess.SubprocessError, OSError) so programming errors in the NM
  restart block are no longer silently swallowed

- api_v3/execute_system_action: add explicit subprocess.TimeoutExpired
  handler before the generic Exception catch; returns action-specific
  message with 'status','message','returncode','stdout','stderr' fields
  so the UI receives a precise, actionable payload instead of the generic
  'Failed to execute system action' string

- plugins_manager.js: move searchPluginStore into .finally() so the
  plugin store renders regardless of whether loadInstalledPlugins succeeds
  or fails; .catch() still logs the error

- first_time_install.sh: add safe_plugin_rm.sh NOPASSWD rule to the
  /tmp/ledmatrix_web_sudoers block; configure_web_sudo.sh had this rule
  but the standalone installer never granted it, leaving plugin removal
  broken after first-time install

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* refactor(api): resolve sudo/systemctl/reboot/poweroff paths at startup

Use shutil.which() with safe fallbacks for the four privileged binaries
instead of relying on bare names being resolved by the subprocess shell
search. Resolves paths once at module load rather than per-call.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

---------

Co-authored-by: Chuck <chuck@example.com>
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
 2026-05-12 17:58:51 -04:00
ChuckClaude Sonnet 4.6Chuck
 		ceb4c4105f fix(wifi): reliable open AP with captive portal — tested on Trixie Pi (#320) 
* fix(wifi): create truly open AP via nmcli connection add; add captive portal to nmcli path

nmcli device wifi hotspot always attaches a WPA2 PSK on Bookworm/Trixie
and silently ignores post-creation security modifications, causing users
to be prompted for an unknown password. Switch to nmcli connection add
with 802-11-wireless.mode ap and no security section — NM cannot auto-add
a password to a profile that has no 802-11-wireless-security block.

Also:
- Remove dead DEFAULT_AP_PASSWORD / ap_password config field (stored but
  never passed to hostapd or nmcli, causing user confusion)
- Add iptables port 80→5000 redirect to the nmcli AP path so captive portal
  auto-popup works on phones without hostapd (previously only worked on
  the hostapd path)
- Clean up iptables rules on disable for the nmcli path
- Improve LED message on AP enable: show SSID, "No password", and IP:port
  on both paths so users know exactly how to connect
- Fix systemd template: replace hardcoded /home/ledpi/LEDMatrix/ with
  __PROJECT_ROOT_DIR__ placeholder (install script already writes correct path)

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix(wifi): address Codacy review findings in AP mode implementation

- Validate ap_ssid/ap_channel from config before passing to subprocess
  (printable ASCII ≤32 chars; channel 1-14) to prevent command injection

- Fix INPUT iptables rule: PREROUTING redirects port 80→5000 so the INPUT
  chain sees dport=5000, not 80. Old INPUT rule on port 80 was a no-op.

- Refactor iptables setup/teardown into _setup_iptables_redirect() and
  _teardown_iptables_redirect() helpers, eliminating duplicate logic in
  the hostapd and nmcli paths

- Save/restore ip_forward state (via /tmp/ledmatrix_ip_forward_saved)
  instead of forcing it to 0 on cleanup, which could break VPNs or
  bridges already relying on forwarding

- nmcli path skips ip_forward management entirely: NM's ipv4.method=shared
  already manages it for the duration of the connection

- Fix _get_ap_status_nmcli() verification: new 'connection add type wifi'
  profiles have type '802-11-wireless', not 'hotspot', so verification was
  always returning False. Now also matches by our known connection name.

- Remove SSID-based connection deletion: deleting any profile whose SSID
  matched the AP SSID could destroy a user's saved home WiFi profile.
  Now only deletes by our application-managed profile names.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix(plugins): fix async race in refreshPlugins; use cache TTL to gate re-swap metadata fetch

refreshPlugins() called searchPluginStore(true) and showNotification() immediately
after refreshInstalledPlugins() without awaiting the returned Promise, so
window.installedPlugins could still be stale when the store rendered its
Installed/Reinstall badges. Chain .then() so both run only after the fetch
completes.

In initializePlugins(), the re-swap path always passed fetchCommitInfo=false to
searchPluginStore, skipping GitHub metadata even when the 5-minute cache TTL had
expired. Add storeCacheExpired() helper and compute isReswapWarm = _reswap &&
!storeCacheExpired() so fresh metadata is fetched whenever the cache is cold,
regardless of whether the render is a first load or a tab re-swap.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix: address three wifi_manager and one plugins_manager review findings

wifi_manager.py:
- _create_hostapd_config: use _validate_ap_config() for ssid/channel instead
  of raw self.config values; strip newlines from SSID to prevent config-file
  injection via the generated hostapd.conf
- _setup_iptables_redirect: check return codes of sysctl ip_forward enable and
  both iptables -A calls; on any failure log the error output, call
  _teardown_iptables_redirect() to restore state, and return False instead of
  silently succeeding
- _enable_ap_mode_nmcli_hotspot: on AP verification failure roll back fully —
  tear down iptables redirect, delete the LEDMatrix-Setup-AP connection profile,
  clear the LED message — before returning False

plugins_manager.js:
- initializePlugins: chain searchPluginStore(!isReswapWarm) inside
  loadInstalledPlugins().then() so window.installedPlugins is populated before
  the store renders Installed/Reinstall badges (same pattern applied to
  refreshPlugins() in the previous commit)

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix(wifi): use _find_command_path for iptables/sysctl; harden ip_forward save/restore

Add _find_command_path() helper that extends _check_command()'s sbin-aware lookup to
return the absolute binary path rather than a boolean. Use it in
_setup_iptables_redirect and _teardown_iptables_redirect so iptables and sysctl are
resolved via /sbin or /usr/sbin even when those directories are absent from PATH in
systemd service environments.

Also harden the ip_forward save/restore logic:
- Read ip_forward from /proc/sys/net/ipv4/ip_forward (no subprocess, no PATH
  dependency) instead of spawning sysctl -n
- Skip the sysctl -w ip_forward=1 write when the value is already "1" to avoid
  mutating state owned by another service (VPN, NM shared mode, bridge)
- Track save success via presence of the save file: if the /proc read or file write
  fails, leave the file absent so teardown knows not to restore
- In _teardown_iptables_redirect, only restore ip_forward when the save file exists;
  if absent, leave the current value untouched rather than forcing "0"

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix(wifi): check _setup_iptables_redirect return; fix hostapd LED SSID; teardown on exception

- Both AP startup paths (hostapd and nmcli) now check the bool returned by
  _setup_iptables_redirect() and treat False as a hard failure: the hostapd
  path stops hostapd/dnsmasq and returns an error tuple; the nmcli path brings
  down and deletes the LEDMatrix-Setup-AP profile and clears the LED message

- _enable_ap_mode_hostapd's LED message now calls _validate_ap_config() to get
  the same sanitized SSID that _create_hostapd_config() uses, so the displayed
  name always matches the AP actually broadcast by hostapd

- _setup_iptables_redirect's outer except block now calls
  _teardown_iptables_redirect() before returning False so partial iptables/
  ip_forward state is always cleaned up on unexpected exceptions; cleanup
  exceptions are caught and logged separately

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* test(wifi): add unit tests for AP mode — open network, iptables, LED, cleanup ordering

Six pytest unit tests covering the five review scenarios. All subprocess and
filesystem side-effects are mocked so the tests run without root, hardware, or
a Pi OS environment.

1. test_nmcli_ap_profile_has_no_security_params — asserts the nmcli connection
   add command has no key-mgmt / psk / WPA arguments and sets mode=ap.
2. test_iptables_nat_rules_added_on_ap_start — verifies _setup_iptables_redirect
   emits a PREROUTING REDIRECT 80→5000 rule and an INPUT ACCEPT rule for port
   5000 (not 80, which never hits INPUT after PREROUTING rewrites it).
3. test_iptables_rules_and_ip_forward_reverted_on_teardown — verifies the -D
   PREROUTING/-D INPUT calls and that sysctl restores the saved ip_forward value
   and removes the save file.
4. test_ip_forward_not_restored_when_save_file_absent — verifies teardown skips
   sysctl when the save file was never written, preventing blind ip_forward=0 on
   systems using ip_forward for VPNs or NM shared mode.
5. test_led_message_shows_ssid_no_password_and_url — asserts the LED message
   includes the SSID, 'No password', and the 192.168.4.1:5000 setup URL.
6. test_existing_ap_profiles_deleted_before_new_profile_created — asserts all
   known profile names are targeted for deletion before 'nmcli connection add'.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* feat(wifi): adopt adsb-feeder-image hotspot patterns — DNS spoofing, connectivity check, idle timeout, wrong-password UX, watchdog escalation

Inspired by the production-proven approach in dirkhh/adsb-feeder-image.

1. DNS spoofing for automatic captive-portal popup (Change 1 — Critical)
   Write /etc/NetworkManager/dnsmasq-shared.d/ledmatrix-captive.conf with
   address=/#/192.168.4.1 before nmcli connection up so NM's built-in
   dnsmasq (ipv4.method=shared) resolves every hostname to the AP IP.
   This triggers the OS captive-portal popup automatically on iOS / Android /
   Windows / macOS — no manual navigation to 192.168.4.1:5000/setup required.
   New helpers: _write_nm_dnsmasq_captive_conf / _remove_nm_dnsmasq_captive_conf.
   New constants: NM_DNSMASQ_SHARED_DIR / NM_DNSMASQ_SHARED_CONF.

2. Real internet connectivity check (Change 2 — High)
   Add _check_internet_connectivity() (ping 8.8.8.8 + HTTP fallback).
   check_and_manage_ap_mode() now considers a device "disconnected" when nmcli
   shows connected but no real internet reachability, matching adsb-feeder's
   multi-method gateway/DNS/HTTP test approach.

3. AP idle timeout (Change 3 — Medium)
   Track _ap_enabled_at timestamp in enable_ap_mode(). Add _has_ap_clients()
   using 'iw dev <iface> station dump'. check_and_manage_ap_mode() auto-disables
   AP after ap_idle_timeout_minutes (default 15) with no associated clients.

4. Wrong-password error feedback (Change 4 — Medium)
   _connect_nmcli() detects "Secrets were required" / "authentication rejected"
   in nmcli stderr and prefixes the message with "wrong_password: ".
   The /api/v3/wifi/connect route propagates error_type="wrong_password" in the
   JSON response. captive_setup.html shows "Incorrect password — try again"
   (keeping the form active) instead of the generic failure message.

5. Escalating watchdog NM restart (Change 5 — Low)
   wifi_monitor_daemon.py tracks _consecutive_internet_failures. After
   _nm_restart_threshold (5) consecutive checks where nmcli shows connected but
   internet is unreachable, restart NetworkManager as a recovery step.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix(wifi): restore safe AP-enable trigger; decouple internet check from AP logic

The previous commit introduced _check_internet_connectivity() into
check_and_manage_ap_mode(), which shared the same _disconnected_checks counter
that triggers AP enable. This created a false-positive risk: 90 seconds of
packet loss on working WiFi would enable AP mode and kick off the connection.

Fix: restore nmcli association state as the sole AP-enable trigger (original,
safe behaviour). The internet connectivity check is now used only in the daemon
watchdog for the NM-restart escalation — matching how adsb-feeder-image actually
structures the two concerns (initial setup detection vs. ongoing monitoring).

Also clarify daemon comment: the connectivity check runs once per cycle in the
watchdog block, not inside check_and_manage_ap_mode, so there is no double-call.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix(wifi): remove PMF setting from open AP profile — breaks nmcli connection add on Trixie NM 1.52+

802-11-wireless-security.pmf is only valid within a security section that also
includes key-mgmt. Adding it to an open-network profile causes NM 1.52+ to
reject the connection add with 'key-mgmt: property is missing'. PMF has no
meaning for open APs (it only applies to WPA2/WPA3), so the setting is simply
removed rather than worked around.

Found by testing on devpi (Trixie, NM 1.52.1).

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix(wifi): add nftables fallback for port redirect; graceful degradation when neither available

Tested on devpi (Trixie, NM 1.52.1): iptables is not installed; nftables is.
The original code called _setup_iptables_redirect() and treated 'iptables not
found' as a hard failure, rolling back the entire AP setup.

Changes:
- _setup_iptables_redirect() now tries iptables first, then nftables as a
  fallback. When neither is available it logs a warning and returns True so
  the AP still comes up (DNS spoofing still triggers the captive portal popup;
  users land on port 5000 directly instead of being auto-redirected from 80).
- Split into _setup_iptables_redirect_iptables() and
  _setup_iptables_redirect_nftables() for clarity.
- Added _redirect_backend instance var ("iptables" | "nftables" | None) so
  _teardown_iptables_redirect() uses the same tool that setup used.
- nftables teardown: deletes the 'ledmatrix' table (clean, no leftover rules).
- iptables teardown: unchanged logic (ip_forward save/restore).
- Also removed the PMF workaround for Trixie: 802-11-wireless-security.pmf
  requires key-mgmt to also be set, breaking open-network creation on NM 1.52+.
  Open APs have no management frame protection by definition.
- Update teardown test to set _redirect_backend = "iptables" before calling it.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix(wifi): public check_internet_connectivity(); absolute systemctl path; stricter mode assertion

wifi_manager.py:
- Add public check_internet_connectivity() wrapping the private method so the
  daemon does not reach into the private API

wifi_monitor_daemon.py:
- Call wifi_manager.check_internet_connectivity() instead of the private
  _check_internet_connectivity()
- Use /usr/bin/systemctl (absolute path) instead of bare "systemctl"
- Wrap NM restart in try/except with check=True; only reset
  _consecutive_internet_failures on success — on CalledProcessError or other
  exception, log the error and leave the counter unchanged so the next cycle
  retries

test/test_wifi_manager_ap.py:
- Replace loose `assert "ap" in add_calls[0]` (list-membership check that
  could be satisfied by any element equal to "ap") with an explicit key/value
  check: locate "802-11-wireless.mode" in the command list and assert the next
  element is exactly "ap"

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

---------

Co-authored-by: Chuck <chuck@example.com>
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
 2026-05-03 08:25:20 -04:00
ChuckClaude Opus 4.6Chuck
 		781224591f fix: post-audit follow-up code fixes (cache, fonts, icons, dev script) (#307) 
* fix: post-audit follow-up code fixes (cache, fonts, icons, dev script, CI)

The docs refresh effort (#306, ledmatrix-plugins#92) surfaced seven
code bugs that were intentionally left out of the docs PRs because
they required code changes rather than doc fixes. This PR addresses
the six that belong in LEDMatrix (the seventh — a lacrosse-scoreboard
mode rename — lives in the plugins repo).

Bug 1: cache_manager.delete() AttributeError
  src/common/api_helper.py:287 and
  src/plugin_system/resource_monitor.py:343 both call
  cache_manager.delete(key), which doesn't exist — only
  clear_cache(key=None). Added a delete() alias method on
  CacheManager that forwards to clear_cache(key). Reverts the
  "There is no delete() method" wording in DEVELOPER_QUICK_REFERENCE,
  .cursorrules so the docs match the new shim.

Bug 2: dev_plugin_setup.sh PROJECT_ROOT resolution
  scripts/dev/dev_plugin_setup.sh:9 set PROJECT_ROOT to SCRIPT_DIR
  instead of walking up two levels to the repo root, so PLUGINS_DIR
  resolved to scripts/dev/plugins/ and created symlinks under the
  script's own directory. Fixed the path and removed the stray
  scripts/dev/plugins/of-the-day symlink left by earlier runs.

Bug 3: plugin custom icons regressed from v2 to v3
  web_interface/blueprints/api_v3.py built the /plugins/installed
  response without including the manifest's "icon" field, and
  web_interface/templates/v3/base.html hardcoded
  fas fa-puzzle-piece in all three plugin-tab render sites. Pass
  the icon through the API and read it from the templates with a
  puzzle-piece fallback. Reverts the "currently broken" banners in
  docs/PLUGIN_CUSTOM_ICONS.md and docs/PLUGIN_CUSTOM_ICONS_FEATURE.md.

Bug 4: register_plugin_fonts was never wired up
  src/font_manager.py:150 defines register_plugin_fonts(plugin_id,
  font_manifest) but nothing called it, so plugin manifests with a
  "fonts" block were silently no-ops. Wired the call into
  PluginManager.load_plugin() right after plugin_loader.load_plugin
  returns. Reverts the "not currently wired" warning in
  docs/FONT_MANAGER.md's "For Plugin Developers" section.

Bug 5: dead web_interface_v2 import pattern (LEDMatrix half)
  src/base_odds_manager.py had a try/except importing
  web_interface_v2.increment_api_counter, falling back to a no-op
  stub. The module doesn't exist anywhere in the v3 codebase and
  no API metrics dashboard reads it. Deleted the import block and
  the single call site; the plugins-repo half of this cleanup lands
  in ledmatrix-plugins#<next>.

Bug 7: no CI test workflow
  .github/workflows/ only contained security-audit.yml; pytest ran
  locally but was not gated on PRs. Added
  .github/workflows/tests.yml running pytest against Python 3.10,
  3.11, 3.12 in EMULATOR=true mode, skipping tests marked hardware
  or slow. Updated docs/HOW_TO_RUN_TESTS.md to reflect that the
  workflow now exists.

Verification done locally:
  - CacheManager.delete(key) round-trips with set/get
  - base_odds_manager imports without the v2 module present
  - dev_plugin_setup.sh PROJECT_ROOT resolves to repo root
  - api_v3 and plugin_manager compile clean
  - tests.yml YAML parses
  - Script syntax check on dev_plugin_setup.sh

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* fix: address CodeRabbit review comments on #307

- src/cache_manager.py: clear_cache(key) treated empty string as
  "wipe all" because of `if key:`. Switched to `key is None`
  branching, made delete(key) and clear_cache(key) reject empty
  strings and None outright with ValueError, and updated both
  docstrings to make the contract explicit. Verified locally
  with a round-trip test that clear_cache() (no arg) still
  wipes everything but clear_cache("") and delete("") raise.

- src/plugin_system/plugin_manager.py: was reaching for the
  font manager via getattr(self.display_manager, 'font_manager',
  None). PluginManager already takes a dedicated font_manager
  parameter (line 54) and stores it as self.font_manager
  (line 69), so the old path was both wrong and could miss the
  font manager entirely when the host injects them separately.
  Switched to self.font_manager directly with the same try/except
  warning behavior.

- web_interface/templates/v3/base.html: in the full plugin-tab
  renderer, the icon was injected with
  `<i class="${escapeHtml(plugin.icon)}">` — but escapeHtml only
  escapes <, >, and &, not double quotes, so a manifest with a
  quote in its icon string could break out of the class
  attribute. Replaced the innerHTML template with createElement
  for the <i> tag, set className from plugin.icon directly
  (no string interpolation), and used a text node for the
  label. Same fix shape would also harden the two stub-renderer
  sites at line 515 / 774, but those already escape `"` to
  &quot; and CodeRabbit only flagged this site, so leaving them
  for now.

- docs/FONT_MANAGER.md: clarified that the Manual Font Overrides
  *workflow* (set_override / remove_override / font_overrides.json)
  is the supported override path today, and only the Fonts tab
  in the web UI is the placeholder. Previous wording conflated
  the two and made it sound like overrides themselves were
  broken.

- docs/HOW_TO_RUN_TESTS.md: replaced the vague "see the PR
  adding it" with a concrete link to #307 and a note that the
  workflow file itself is held back pending the workflow scope.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

---------

Co-authored-by: Chuck <chuck@example.com>
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-04-08 09:25:12 -04:00
ChuckClaude Opus 4.6Chuck
 		6812dfe7a6 docs: refresh and correct stale documentation across repo (#306) 
* docs: refresh and correct stale documentation across repo

Walked the README and docs/ tree against current code and fixed several
real bugs and many stale references. Highlights:

User-facing
- README.md: web interface install instructions referenced
  install_web_service.sh at the repo root, but it actually lives at
  scripts/install/install_web_service.sh.
- docs/GETTING_STARTED.md: every web UI port reference said 5050, but
  the real server in web_interface/start.py:123 binds 5000. Same bug
  was duplicated in docs/TROUBLESHOOTING.md (17 occurrences). Fixed
  both.
- docs/GETTING_STARTED.md: rewrote tab-by-tab instructions. The doc
  referenced "Plugin Store", "Plugin Management", "Sports Configuration",
  "Durations", and "Font Management" tabs - none of which exist. Real
  tabs (verified in web_interface/templates/v3/base.html) are: Overview,
  General, WiFi, Schedule, Display, Config Editor, Fonts, Logs, Cache,
  Operation History, Plugin Manager (+ per-plugin tabs).
- docs/GETTING_STARTED.md: removed references to a "Test Display"
  button (doesn't exist) and "Show Now" / "Stop" plugin buttons. Real
  controls are "Run On-Demand" / "Stop On-Demand" inside each plugin's
  tab (partials/plugin_config.html:792).
- docs/TROUBLESHOOTING.md: removed dead reference to
  troubleshoot_weather.sh (doesn't exist anywhere in the repo); weather
  is now a plugin in ledmatrix-plugins.

Developer-facing
- docs/PLUGIN_API_REFERENCE.md: documented draw_image() doesn't exist
  on DisplayManager. Real plugins paste onto display_manager.image
  directly (verified in src/base_classes/{baseball,basketball,football,
  hockey}.py). Replaced with the canonical pattern.
- docs/PLUGIN_API_REFERENCE.md: documented cache_manager.delete() doesn't
  exist. Real method is clear_cache(key=None). Updated the section.
- docs/PLUGIN_API_REFERENCE.md: added 10 missing BasePlugin methods that
  the doc never mentioned: dynamic-duration hooks, live-priority hooks,
  and the full Vegas-mode interface.
- docs/PLUGIN_DEVELOPMENT_GUIDE.md: same draw_image fix.
- docs/DEVELOPMENT.md: corrected the "Plugin Submodules" section. Plugins
  are NOT git submodules - .gitmodules only contains
  rpi-rgb-led-matrix-master. Plugins are installed at runtime into the
  plugins directory configured by plugin_system.plugins_directory
  (default plugin-repos/). Both internal links in this doc were also
  broken (missing relative path adjustment).
- docs/HOW_TO_RUN_TESTS.md: removed pytest-timeout from install line
  (not in requirements.txt) and corrected the test/integration/ path
  (real integration tests are at test/web_interface/integration/).
  Replaced the fictional file structure diagram with the real one.
- docs/EMULATOR_SETUP_GUIDE.md: clone URL was a placeholder; default
  pixel_size was documented as 16 but emulator_config.json ships with 5.

Index
- docs/README.md: rewrote. Old index claimed "16-17 files after
  consolidation" but docs/ actually has 38 .md files. Four were missing
  from the index entirely (CONFIG_DEBUGGING, DEV_PREVIEW,
  PLUGIN_ERROR_HANDLING, STARLARK_APPS_GUIDE). Trimmed the navel-gazing
  consolidation/statistics sections.

Out of scope but worth flagging:
- src/plugin_system/resource_monitor.py:343 and src/common/api_helper.py:287
  call cache_manager.delete(key) but no such method exists on
  CacheManager. Both call sites would AttributeError at runtime if hit.
  Not fixed in this docs PR - either add a delete() shim or convert
  callers to clear_cache().

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* docs: fix WEB_INTERFACE_GUIDE and WIFI_NETWORK_SETUP

WEB_INTERFACE_GUIDE.md
- Web UI port: 5050 -> 5000 (4 occurrences)
- Tab list was almost entirely fictional. Documented tabs:
  General Settings, Display Settings, Durations, Sports Configuration,
  Plugin Management, Plugin Store, Font Management. None of these
  exist. Real tabs (verified in web_interface/templates/v3/base.html:
  935-1000): Overview, General, WiFi, Schedule, Display, Config Editor,
  Fonts, Logs, Cache, Operation History, plus Plugin Manager and
  per-plugin tabs in the second nav row. Rewrote the navigation
  section, the General/Display/Plugin sections, and the Common Tasks
  walkthroughs to match.
- Quick Actions list referenced "Test Display" button (doesn't exist).
  Replaced with the real button list verified in
  partials/overview.html:88-152: Start/Stop Display, Restart Display
  Service, Restart Web Service, Update Code, Reboot, Shutdown.
- API endpoints used /api/* paths. The api_v3 blueprint mounts at
  /api/v3 (web_interface/app.py:144), so the real paths are
  /api/v3/config/main, /api/v3/system/status, etc. Fixed.
- Removed bogus "Sports Configuration tab" walkthrough; sports
  favorites live inside each scoreboard plugin's own tab now.
- Plugin directory listed as /plugins/. Real default is plugin-repos/
  (verified in config/config.template.json:130 and
  display_controller.py:132); plugins/ is a fallback.
- Removed "Swipe navigation between tabs" mobile claim (not implemented).

WIFI_NETWORK_SETUP.md
- 21 occurrences of port 5050 -> 5000.
- All /api/wifi/* curl examples used the wrong path. The real wifi
  API routes are at /api/v3/wifi/* (api_v3.py:6367-6609). Fixed.
- ap_password default was documented as "" (empty/open network) but
  config/wifi_config.json ships with "ledmatrix123". Updated the
  Quick Start, Configuration table, AP Mode Settings section, and
  Security Recommendations to match. Also clarified that setting
  ap_password to "" is the way to make it an open network.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* docs: fix ADVANCED_FEATURES and REST_API_REFERENCE

REST_API_REFERENCE.md
- Wrong path: /fonts/delete/<font_family> -> /fonts/<font_family>
  (verified the real DELETE route in
  web_interface/blueprints/api_v3.py).
- Diffed the documented routes against the real api_v3 blueprint
  (92 routes vs the 71 documented). Added missing sections:
  - Error tracking (/errors/summary, /errors/plugin/<id>, /errors/clear)
  - Health (/health)
  - Schedule dim/power (/config/dim-schedule GET/POST)
  - Plugin-specific endpoints (calendar/list-calendars,
    of-the-day/json/upload+delete, plugins/<id>/static/<path>)
  - Starlark Apps (12 endpoints: status, install-pixlet, apps CRUD,
    repository browse/install, upload)
  - Font preview (/fonts/preview)
- Updated table of contents with the new sections.
- Added a footer note that the API blueprint mounts at /api/v3
  (app.py:144) and that SSE stream endpoints are defined directly on
  the Flask app at app.py:607-615.

ADVANCED_FEATURES.md
- Vegas Scroll Mode section was actually accurate (verified all
  config keys match src/vegas_mode/config.py:15-30).

- On-Demand Display section had multiple bugs:
  - 5 occurrences of port 5050 -> 5000
  - All API paths missing /v3 (e.g. /api/display/on-demand/start
    should be /api/v3/display/on-demand/start)
  - "Settings -> Plugin Management -> Show Now Button" UI flow doesn't
    exist. Real flow: open the plugin's tab in the second nav row,
    click Run On-Demand / Stop On-Demand.
  - "Python API Methods" section showed
    controller.show_on_demand() / clear_on_demand() /
    is_on_demand_active() / get_on_demand_info() — none of these
    methods exist on DisplayController. The on-demand machinery is
    all internal (_set_on_demand_*, _activate_on_demand, etc) and
    is driven through the cache_manager. Replaced the section with
    a note pointing to the REST API.
  - All Use Case Examples used the same fictional Python calls.
    Replaced with curl examples against the real API.

- Cache Management section claimed "On-demand display uses Redis cache
  keys". LEDMatrix doesn't use Redis — verified with grep that
  src/cache_manager.py has no redis import. The cache is file-based,
  managed by CacheManager (file at /var/cache/ledmatrix/ or fallback
  paths). Rewrote the manual recovery section:
  - Removed redis-cli commands
  - Replaced cache.delete() Python calls with cache.clear_cache()
    (the real public method per the same bug already flagged in
    PLUGIN_API_REFERENCE.md)
  - Replaced "Settings -> Cache Management" with the real Cache tab
  - Documented the actual cache directory candidates

- Background Data Service section:
  - Used "nfl_scoreboard" as the plugin id in the example.
    The real plugin is "football-scoreboard" (handles both NFL and
    NCAA). Fixed.
  - "Implementation Status: Phase 1 NFL only / Phase 2 planned"
    section was severely outdated. The background service is now
    used by all sports scoreboards (football, hockey, baseball,
    basketball, soccer, lacrosse, F1, UFC), the odds ticker, and
    the leaderboard plugin. Replaced with a current "Plugins using
    the background service" note.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* docs: fix plugin config + store + dependency docs

PLUGIN_STORE_GUIDE.md
- 19 occurrences of port 5050 -> 5000
- All API paths missing /v3 (e.g. /api/plugins/install ->
  /api/v3/plugins/install). Bulk fix.

PLUGIN_REGISTRY_SETUP_GUIDE.md
- Same port + /api/v3 fixes (3 occurrences each)
- "Go to Plugin Store tab" -> "Open the Plugin Manager tab and scroll
  to the Install from GitHub section" (the real flow for registry
  setup is the GitHub install section, not the Plugin Store search)

PLUGIN_CONFIG_QUICK_START.md
- Port 5001 -> 5000 (5001 is the dev_server.py default, not the web UI)
- "Plugin Store tab" install flow -> real Plugin Manager + Plugin Store
  section + per-plugin tab in second nav row
- Removed reference to PLUGIN_CONFIG_TABS_SUMMARY.md (archived doc)

PLUGIN_CONFIGURATION_TABS.md
- "Plugin Management vs Configuration" section confusingly described
  a "Plugins Tab" that doesn't exist as a single thing. Rewrote to
  describe the real two-piece structure: Plugin Manager tab (browse,
  install, toggle) vs per-plugin tabs (configure individual plugins).

PLUGIN_DEPENDENCY_GUIDE.md
- Port 5001 -> 5000

PLUGIN_DEPENDENCY_TROUBLESHOOTING.md
- Wrong port (8080) and wrong UI nav ("Plugin Store or Plugin
  Management"). Fixed to the real flow.

PLUGIN_QUICK_REFERENCE.md
- "Plugin Location: ./plugins/ directory" -> default is plugin-repos/
  (verified in config/config.template.json:130 and
  display_controller.py:132). plugins/ is a fallback.
- File structure diagram showed plugins/ -> plugin-repos/.
- Web UI install flow: "Plugin Store tab" -> "Plugin Manager tab ->
  Plugin Store section". Also fixed Configure ⚙️ button (doesn't
  exist) and "Drag and drop reorder" (not implemented).
- API examples: replaced ad-hoc Python pseudocode with real curl
  examples against /api/v3/plugins/* endpoints. Pointed at
  REST_API_REFERENCE.md for the full list.
- "Migration Path Phase 1-5" was a roadmap written before the plugin
  system shipped. The plugin system is now stable and live. Removed
  the migration phases as they're history, not a roadmap.
- "Quick Migration" section called scripts/migrate_to_plugins.py
  which doesn't exist anywhere in the repo. Removed.
- "Plugin Registry Structure" referenced
  ChuckBuilds/ledmatrix-plugin-registry which doesn't exist. The
  real registry is ChuckBuilds/ledmatrix-plugins. Fixed.
- "Next Steps" / "Questions to Resolve" sections were
  pre-implementation planning notes. Replaced with a "Known
  Limitations" section that documents the actually-real gaps
  (sandboxing, resource limits, ratings, auto-updates).

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* docs: fix misc remaining docs (architecture, dev quickref, sub-dir READMEs)

PLUGIN_ARCHITECTURE_SPEC.md
- Added a banner at the top noting this is a historical design doc
  written before the plugin system shipped. The doc is ~1900 lines
  with 13 stale /api/plugins/* paths (real is /api/v3/plugins/*),
  references to web_interface_v2.py (current is app.py), and a
  Migration Strategy / Implementation Roadmap that's now history.
  Banner points readers at the current docs
  (PLUGIN_DEVELOPMENT_GUIDE, PLUGIN_API_REFERENCE,
  REST_API_REFERENCE) without needing to retrofit every section.

PLUGIN_CONFIG_ARCHITECTURE.md
- 10 occurrences of /api/plugins/* missing /v3 prefix. Bulk fixed.

DEVELOPER_QUICK_REFERENCE.md
- cache_manager.delete("key") -> cache_manager.clear_cache("key")
  with comment noting delete() doesn't exist. Same bug already
  documented in PLUGIN_API_REFERENCE.md.

SSH_UNAVAILABLE_AFTER_INSTALL.md
- 4 occurrences of port 5001 -> 5000 in AP-mode and Ethernet/WiFi
  recovery instructions.

PLUGIN_CUSTOM_ICONS_FEATURE.md
- Port 5001 -> 5000.

CONFIG_DEBUGGING.md
- Documented /api/v3/config/plugin/<id> and /api/v3/config/validate
  endpoints don't exist. Replaced with the real endpoints:
  /api/v3/config/main, /api/v3/plugins/schema?plugin_id=,
  /api/v3/plugins/config?plugin_id=. Added a note that validation
  runs server-side automatically on POST.

STARLARK_APPS_GUIDE.md
- "Plugins -> Starlark Apps" UI navigation path doesn't exist (5
  occurrences). Replaced with the real path: Plugin Manager tab,
  then the per-plugin Starlark Apps tab in the second nav row.
- "Navigate to Plugins" install step -> Plugin Manager tab.

web_interface/README.md
- Documented several endpoints that don't exist in the api_v3
  blueprint:
  - GET /api/v3/plugins (list) -> /api/v3/plugins/installed
  - GET /api/v3/plugins/<id> -> doesn't exist
  - POST /api/v3/plugins/<id>/config -> POST /api/v3/plugins/config
  - GET /api/v3/plugins/<id>/enable + /disable -> POST /api/v3/plugins/toggle
  - GET /api/v3/store/plugins -> /api/v3/plugins/store/list
  - POST /api/v3/store/install/<id> -> POST /api/v3/plugins/install
  - POST /api/v3/store/uninstall/<id> -> POST /api/v3/plugins/uninstall
  - POST /api/v3/store/update/<id> -> POST /api/v3/plugins/update
  - POST /api/v3/display/start/stop/restart -> POST /api/v3/system/action
  - GET /api/v3/display/status -> GET /api/v3/system/status
- Also fixed config/secrets.json -> config/config_secrets.json
- Replaced the per-section endpoint duplication with a current real
  endpoint list and a pointer to docs/REST_API_REFERENCE.md.
- Documented that SSE stream endpoints are defined directly on the
  Flask app at app.py:607-615, not in the api_v3 blueprint.

scripts/install/README.md
- Was missing 3 of the 9 install scripts in the directory:
  one-shot-install.sh, configure_wifi_permissions.sh, and
  debug_install.sh. Added them with brief descriptions.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* docs: clarify plugin paths and fix systemd manual install bug

PLUGIN_DEVELOPMENT_GUIDE.md
- Added a "Plugin directory note" callout near the top explaining
  the plugins/ vs plugin-repos/ split:
  - Dev workflow uses plugins/ (where dev_plugin_setup.sh creates
    symlinks)
  - Production / Plugin Store uses plugin-repos/ (the configurable
    default per config.template.json:130)
  - The plugin loader falls back to plugins/ so dev symlinks are
    picked up automatically (schema_manager.py:77)
  - User can set plugins_directory to "plugins" in the General tab
    if they want both to share a directory

CLAUDE.md
- The Project Structure section had plugins/ and plugin-repos/
  exactly reversed:
  - Old: "plugins/ - Installed plugins directory (gitignored)"
         "plugin-repos/ - Development symlinks to monorepo plugin dirs"
  - Real: plugin-repos/ is the canonical Plugin Store install
    location and is not gitignored. plugins/* IS gitignored
    (verified in .gitignore) and is the legacy/dev location used by
    scripts/dev/dev_plugin_setup.sh.
  Reversed the descriptions and added line refs.

systemd/README.md
- "Manual Installation" section told users to copy the unit file
  directly to /etc/systemd/system/. Verified the unit file in
  systemd/ledmatrix.service contains __PROJECT_ROOT_DIR__
  placeholders that the install scripts substitute at install time.
  A user following the manual steps would get a service that fails
  to start with "WorkingDirectory=__PROJECT_ROOT_DIR__" errors.
  Added a clear warning and a sed snippet that substitutes the
  placeholder before installing.

src/common/README.md
- Was missing 2 of the 11 utility modules in the directory
  (verified with ls): permission_utils.py and cli.py. Added brief
  descriptions for both.

Out-of-scope code bug found while auditing (flagged but not fixed):
- scripts/dev/dev_plugin_setup.sh:9 sets PROJECT_ROOT="$SCRIPT_DIR"
  which resolves to scripts/dev/, not the project root. This means
  the script's PLUGINS_DIR resolves to scripts/dev/plugins/ instead
  of the project's plugins/ — confirmed by the existence of
  scripts/dev/plugins/of-the-day/ from prior runs. Real fix is to
  set PROJECT_ROOT="$(cd "$SCRIPT_DIR/../.." && pwd)". Not fixing in
  this docs PR.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* docs: flag aspirational/regressed features in plugin docs

These docs describe features that exist as documented in the doc but
either never wired up or regressed when v3 shipped. Each gets a clear
status banner so plugin authors don't waste time chasing features that
don't actually work.

FONT_MANAGER.md
- The "For Plugin Developers / Plugin Font Registration" section
  documents adding a "fonts" block to manifest.json that gets
  registered via FontManager.register_plugin_fonts(). The method
  exists at src/font_manager.py:150 but is **never called from
  anywhere** in the codebase (verified: zero callers). A plugin
  shipping a manifest "fonts" block has its fonts silently ignored.
  Added a status warning and a note about how to actually ship plugin
  fonts (regular files in the plugin dir, loaded directly).

PLUGIN_IMPLEMENTATION_SUMMARY.md
- Added a top-level status banner.
- Architecture diagram referenced src/plugin_system/registry_manager.py
  (which doesn't exist) and listed plugins/ as the install location.
  Replaced with the real file list (plugin_loader, schema_manager,
  health_monitor, operation_queue, state_manager) and pointed at
  plugin-repos/ as the default install location.
- "Dependency Management: Virtual Environments" — verified there's no
  per-plugin venv. Removed the bullet and added a note that plugin
  Python deps install into the system Python environment, with no
  conflict resolution.
- "Permission System: File Access Control / Network Access /
  Resource Limits / CPU and memory constraints" — none of these
  exist. There's a resource_monitor.py and health_monitor.py for
  metrics/warnings, but no hard caps or sandboxing. Replaced the
  section with what's actually implemented and a clear note that
  plugins run in the same process with full file/network access.

PLUGIN_CUSTOM_ICONS.md and PLUGIN_CUSTOM_ICONS_FEATURE.md
- The custom-icon feature was implemented in the v2 web interface
  via a getPluginIcon() helper in templates/index_v2.html that read
  the manifest "icon" field. When the v3 web interface was built,
  that helper wasn't ported. Verified in
  web_interface/templates/v3/base.html:515 and :774, plugin tab
  icons are hardcoded to `fas fa-puzzle-piece`. The "icon" field in
  plugin manifests is currently silently ignored (verified with grep
  across web_interface/ and src/plugin_system/ — zero non-action-
  related reads of plugin.icon or manifest.icon).
- Added a status banner to both docs noting the regression so plugin
  authors don't think their custom icons are broken in their own
  plugin code.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* docs: fix .cursor/ helper docs

The .cursor/ directory holds the dev-side helper docs that Cursor and
contributors using AI tooling rely on to bootstrap plugin development.
Several of them had the same bug patterns as the user-facing docs.

.cursor/plugin_templates/QUICK_START.md
- "Adding Image Rendering" section showed
  display_manager.draw_image(image, x=0, y=0). That method doesn't
  exist on DisplayManager (same bug as PLUGIN_API_REFERENCE.md and
  PLUGIN_DEVELOPMENT_GUIDE.md). Replaced with the canonical
  display_manager.image.paste((x,y)) pattern, including the
  transparency-mask form.

.cursor/plugins_guide.md
- 10 occurrences of ./dev_plugin_setup.sh — the script lives at
  scripts/dev/dev_plugin_setup.sh, so anyone copy-pasting these
  examples gets "command not found". Bulk fixed via sed.
- "Test with emulator: python run.py --emulator" — there's no
  --emulator flag. Replaced with the real options:
  EMULATOR=true python3 run.py for the full display, or
  scripts/dev_server.py for the dev preview.
- Secrets management section showed a fictional
  "config_secrets": { "api_key": "my-plugin.api_key" } reference
  field. Verified in src/config_manager.py:162-172 that secrets are
  loaded by deep-merging config_secrets.json into the main config.
  There is no separate reference field — just put the secret under
  the same plugin namespace and read it from the merged config.
  Rewrote the section with the real pattern.
- "ssh pi@raspberrypi" -> "ssh ledpi@your-pi-ip" (consistent with
  the rest of LEDMatrix docs which use ledpi as the default user)

.cursor/README.md
- Same ./dev_plugin_setup.sh -> ./scripts/dev/dev_plugin_setup.sh
  fix (×6 occurrences via replace_all).
- Same "python run.py --emulator" -> "EMULATOR=true python3 run.py"
  fix. Also added a pointer to scripts/dev_server.py for previewing
  plugins without running the full display.
- "Example Plugins: plugins/hockey-scoreboard/" — the canonical
  source is the ledmatrix-plugins repo. Installed copies land in
  plugin-repos/ or plugins/. Updated the line to point at the
  ledmatrix-plugins repo and explain both local locations.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* docs: fix .cursorrules — the file Cursor auto-loads to learn the API

This is the file that Cursor reads to learn how plugin development
works. Stale entries here directly mislead AI-assisted plugin authors
on every new plugin. Several of the same bug patterns I've been
fixing in the user-facing docs were here too.

Display Manager section (highest impact)
- "draw_image(image, x, y): Draw PIL Image" — that method doesn't
  exist on DisplayManager. Same bug already fixed in
  PLUGIN_API_REFERENCE.md, PLUGIN_DEVELOPMENT_GUIDE.md,
  ledmatrix-stocks/README.md, and .cursor/plugin_templates/QUICK_START.md.
  Removed the bullet and replaced it with a paragraph explaining the
  real pattern: paste onto display_manager.image directly, then
  update_display(). Includes the transparency-mask form.
- Added the small_font/centered args to draw_text() since they're
  the ones that matter most for new plugin authors
- Added draw_weather_icon since it's commonly used

Cache Manager section
- "delete(key): Remove cached value" — there's no delete() method
  on CacheManager. The real method is clear_cache(key=None) (also
  removes everything when called without args). Same bug as before.
- Added get_cached_data_with_strategy and get_background_cached_data
  since contributors will hit these when working on sports plugins

Plugin System Overview
- "loaded from the plugins/ directory" — clarified that the default
  is plugin-repos/ (per config.template.json:130) with plugins/ as
  the dev fallback used by scripts/dev/dev_plugin_setup.sh

Plugin Development Workflow
- ./dev_plugin_setup.sh -> ./scripts/dev/dev_plugin_setup.sh (×2)
- Manual setup step "Create directory in plugins/<plugin-id>/" ->
  plugin-repos/<plugin-id>/ as the canonical location
- "Use emulator: python run.py --emulator or ./run_emulator.sh"
  — the --emulator flag doesn't exist; ./run_emulator.sh isn't at
  root (it lives at scripts/dev/run_emulator.sh). Replaced with the
  real options: scripts/dev_server.py for dev preview, or
  EMULATOR=true python3 run.py for the full emulator path.

Configuration Management
- "Reference secrets via config_secrets key in main config" — this
  is the same fictional reference syntax I just fixed in
  .cursor/plugins_guide.md. Verified in src/config_manager.py:162-172
  that secrets are deep-merged into the main config; there's no
  separate reference field. Replaced with a clear explanation of
  the deep-merge approach.

Code Organization
- "plugins/<plugin-id>/" -> the canonical location is
  plugin-repos/<plugin-id>/ (or its dev-time symlink in plugins/)
- "see plugins/hockey-scoreboard/ as reference" — the canonical
  source for example plugins is the ledmatrix-plugins repo. Updated
  the pointer.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* Add LICENSE (GPL-3.0) and CONTRIBUTING.md

LICENSE
- The repository previously had no LICENSE file. The README and every
  downstream plugin README already reference GPL-3.0 ("same as
  LEDMatrix project"), but the canonical license text was missing —
  contributors had no formal record of what they were contributing
  under, and GitHub couldn't auto-detect the license for the repo
  banner.
- Added the canonical GPL-3.0 text from
  https://www.gnu.org/licenses/gpl-3.0.txt (verbatim, 674 lines).
- Compatibility verified: rpi-rgb-led-matrix is GPL-2.0-or-later
  (per its COPYING file and README; the "or any later version" clause
  in lib/*.h headers makes GPL-3.0 distribution legal).

CONTRIBUTING.md
- The repository had no CONTRIBUTING file. New contributors had to
  reconstruct the dev setup from DEVELOPMENT.md, PLUGIN_DEVELOPMENT_GUIDE.md,
  SUBMISSION.md, and the root README.
- Added a single page covering: dev environment setup (preview
  server, emulator, hardware), running tests, PR submission flow,
  commit message convention, plugin contribution pointer, and the
  license terms contributors are agreeing to.

> Note for the maintainer: I (the AI assistant doing this audit) am
> selecting GPL-3.0 because every reference in the existing
> documentation already says GPL-3.0 — this commit just makes that
> declaration legally binding by adding the actual file. Please
> confirm during PR review that GPL-3.0 is what you want; if you
> prefer a different license, revert this commit before merging.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* Add CODE_OF_CONDUCT, SECURITY, PR template; link them from README

Tier 1 organizational files that any open-source project at
LEDMatrix's maturity is expected to have. None of these existed
before. They're additive — no existing content was rewritten.

CODE_OF_CONDUCT.md
- Contributor Covenant 2.1 (the de facto standard for open-source
  projects). Mentions both the Discord and the GitHub Security
  Advisories channel for reporting violations.

SECURITY.md
- Private vulnerability disclosure flow with two channels: GitHub
  Security Advisories (preferred) and Discord DM.
- Documents the project's known security model as intentional
  rather than vulnerabilities: no web UI auth, plugins run
  unsandboxed, display service runs as root for GPIO access,
  config_secrets.json is plaintext. These match the limitations
  already called out in PLUGIN_QUICK_REFERENCE.md and the audit
  flagging from earlier in this PR.
- Out-of-scope section points users at upstream
  (rpi-rgb-led-matrix, third-party plugins) so reports land in the
  right place.

.github/PULL_REQUEST_TEMPLATE.md
- 10-line checklist that prompts for the things that would have
  caught the bugs in this very PR: did you load the changed plugin
  once, did you update docs alongside code, are there any plugin
  compatibility implications.
- Linked from CONTRIBUTING.md for the full flow.

README.md
- Added a License section near the bottom (the README previously
  said nothing about the license despite the project being GPL-3.0).
- Added a Contributing section pointing at CONTRIBUTING.md and
  SECURITY.md.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* Customize bug report template for LEDMatrix hardware

The bug_report.md template was the GitHub default and asked
"Desktop (OS/Browser/Version)" and "Smartphone (Device/OS)" — neither
of which is relevant for a project that runs on a Raspberry Pi with
hardware LED panels. A user filing a bug under the old template was
giving us none of the information we'd actually need to triage it.

Replaced with a LEDMatrix-aware template that prompts for:
- Pi model, OS/kernel, panel type, HAT/Bonnet, PWM jumper status,
  display chain dimensions
- LEDMatrix git commit / release tag
- Plugin id and version (if the bug is plugin-related)
- Relevant config snippet (with redaction reminder for API keys)
- journalctl log excerpt with the exact command to capture it
- Optional photo of the actual display for visual issues

Kept feature_request.md as-is — generic content there is fine.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* docs: fix bare /api/plugins paths in PLUGIN_CONFIGURATION_TABS

Found 5 more bare /api/plugins/* paths in PLUGIN_CONFIGURATION_TABS.md
that I missed in the round 2 sweep — they're inside data flow diagrams
and prose ("loaded via /api/plugins/installed", etc.) so the earlier
grep over Markdown code blocks didn't catch them. Fixed all 5 to use
/api/v3/plugins/* (the api_v3 blueprint mount path verified at
web_interface/app.py:144).

Also added a status banner noting that the "Implementation Details"
section references the pre-v3 file layout (web_interface_v2.py,
templates/index_v2.html) which no longer exists. The current
implementation is in web_interface/app.py, blueprints/api_v3.py, and
templates/v3/. Same kind of historical drift I flagged in
PLUGIN_ARCHITECTURE_SPEC.md and the PLUGIN_CUSTOM_ICONS_FEATURE doc.
The user-facing parts of the doc (Overview, Features, Form Generation
Process) are still accurate.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* docs(widgets): list the 20 undocumented built-in widgets

The widget registry README documented 3 widgets (file-upload,
checkbox-group, custom-feeds) but the directory contains 23 registered
widgets total. A plugin author reading this doc would think those 3
were the only built-in options and either reach for a custom widget
unnecessarily or settle for a generic text input.

Verified the actual list with:
  grep -h "register('" web_interface/static/v3/js/widgets/*.js \
    | sed -E "s|.*register\\('([^']+)'.*|\\1|" | sort -u

Added an "Other Built-in Widgets" section after the 3 detailed
sections, listing the remaining 20 with one-line descriptions
organized by category:
- Inputs (6): text-input, textarea, number-input, email-input,
  url-input, password-input
- Selectors (7): select-dropdown, radio-group, toggle-switch,
  slider, color-picker, font-selector, timezone-selector
- Date/time/scheduling (4): date-picker, day-selector, time-range,
  schedule-picker
- Composite/data-source (2): array-table, google-calendar-picker
- Internal (2): notification, base-widget

Pointed at the .js source files as the canonical source for each
widget's exact schema and options — keeps this list low-maintenance
since I'm not duplicating each widget's full options table.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* docs: fix README_NBA_LOGOS and PLUGIN_CONFIGURATION_GUIDE

scripts/README_NBA_LOGOS.md
- "python download_nba_logos.py" — wrong on two counts. The script
  is at scripts/download_nba_logos.py (not the project root), and
  "python" is Python 2 on most systems. Replaced all 4 occurrences
  with "python3 scripts/download_nba_logos.py".
- The doc framed itself as the way to set up "the NBA leaderboard".
  The basketball/leaderboard functionality is now in the
  basketball-scoreboard and ledmatrix-leaderboard plugins (in the
  ledmatrix-plugins repo), which auto-download logos on first run.
  Reframed the script as a pre-population utility for offline / dev
  use cases.
- Bumped the documented Python minimum from 3.7 to 3.9 to match
  the rest of the project.

docs/PLUGIN_CONFIGURATION_GUIDE.md
- The "Plugin Manifest" example was missing 3 fields the plugin
  loader actually requires: id, entry_point, and class_name. A
  contributor copying this manifest verbatim would get
  PluginError("No class_name in manifest") at load time — the same
  loader bug already found in stock-news. Added all three.
- The same example showed config_schema as an inline object. The
  loader expects config_schema to be a file path string (e.g.
  "config_schema.json") with the actual schema in a separate JSON
  file — verified earlier in this audit. Fixed.
- Added a paragraph explaining the loader's required fields and
  the case-sensitivity rule on class_name (the bug that broke
  hello-world's manifest before this PR fixed it).
- "Plugin Manager Class" example had the wrong constructor
  signature: (config, display_manager, cache_manager, font_manager).
  The real BasePlugin.__init__ at base_plugin.py:53-60 takes
  (plugin_id, config, display_manager, cache_manager, plugin_manager).
  A copy-pasted example would TypeError on instantiation. Fixed,
  including a comment noting which attributes BasePlugin sets up.
- Renamed the example class from MyPluginManager to MyPlugin to
  match the project convention (XxxPlugin / XxxScoreboardPlugin
  in actual plugins).

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* docs(requirements): document optional dependencies (scipy, psutil, Flask-Limiter)

A doc-vs-code crosscheck of every Python import in src/ and
web_interface/ against requirements.txt found 3 packages that the
code uses but requirements.txt doesn't list. Verified with grep that
all 3 are wrapped in try/except blocks with documented fallback
paths, so they're optional features rather than missing required
deps:

- scipy           src/common/scroll_helper.py:26
                  → from scipy.ndimage import shift; HAS_SCIPY flag.
                  Used for sub-pixel interpolation in scrolling.
                  Falls back to a simpler shift algorithm without it.

- psutil          src/plugin_system/resource_monitor.py:15
                  → import psutil; PSUTIL_AVAILABLE flag. Used for
                  per-plugin CPU/memory monitoring. Silently no-ops
                  without it.

- flask-limiter   web_interface/app.py:42-43
                  → from flask_limiter import Limiter; wrapped at the
                  caller. Used for accidental-abuse rate limiting on
                  the web interface (not security). Web interface
                  starts without rate limiting when missing.

These were latent in two ways:
1. A user reading requirements.txt thinks they have the full feature
   set after `pip install -r requirements.txt`, but they don't get
   smoother scrolling, plugin resource monitoring, or rate limiting.
2. A contributor who deletes one of the packages from their dev env
   wouldn't know which feature they just lost — the fallbacks are
   silent.

Added an "Optional dependencies" section at the bottom of
requirements.txt with the version constraint, the file:line where
each is used, the feature it enables, and the install command. The
comment-only format means `pip install -r requirements.txt` still
gives the minimal-feature install (preserving current behavior),
while users who want the full feature set can copy the explicit
pip install commands.

Other findings from the same scan that came back as false positives
or known issues:
- web_interface_v2: dead pattern flagged in earlier iteration
  (still no real implementation; affects 11+ plugins via the same
  try/except dead-fallback pattern)
- urllib3: comes with `requests` transitively
- All 'src.', 'web_interface.', 'rgbmatrix', 'RGBMatrixEmulator'
  imports: internal modules
- base_plugin / plugin_manager / store_manager / mocks /
  visual_display_manager: relative imports to local modules
- freetype: false positive (freetype-py is in requirements.txt
  under the package name)

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* docs: fix broken file references found by path-existence crosscheck

Ran a doc-vs-filesystem crosscheck: extracted every backtick-quoted
path with a file extension or known directory prefix from docs/*.md
and verified each exists. After filtering false positives
(placeholder paths, config keys mistaken for paths, paths inside
docs that already have historical-status banners), found 4 real
broken references — 3 fixed in docs, 1 fixed by creating the missing
file:

docs/HOW_TO_RUN_TESTS.md:339
- Claimed ".github/workflows/tests.yml" exists and runs pytest on
  multiple Python versions in CI. There is no such workflow.
  The only GitHub Actions file is security-audit.yml (bandit + semgrep).
- Pytest runs locally but is NOT gated on PRs.
- Replaced the fictional CI section with the actual state and a
  note explaining how someone could contribute a real test workflow.

docs/MIGRATION_GUIDE.md:92
- Referenced scripts/fix_perms/README.md "(if exists)" — the
  hedge betrays that the writer wasn't sure. The README didn't
  exist. The 6 scripts in scripts/fix_perms/ were never documented.
- Created the missing scripts/fix_perms/README.md from scratch
  with one-line descriptions of all 6 scripts (fix_assets,
  fix_cache, fix_plugin, fix_web, fix_nhl_cache, safe_plugin_rm)
  + when-to-use-each guidance + usage examples.
- Updated MIGRATION_GUIDE link to drop the "(if exists)" hedge
  since the file now exists.

docs/FONT_MANAGER.md:376
- "See test/font_manager_example.py for a complete working example"
  — that file does not exist. Verified by listing test/ directory.
- Replaced with a pointer to src/font_manager.py itself and the
  existing scoreboard base classes in src/base_classes/ that
  actually use the font manager API in production.

Path-existence check methodology:
- Walked docs/ recursively, regex-extracted backtick-quoted paths
  matching either /\.(py|sh|json|yml|yaml|md|txt|service|html|js|css|ttf|bdf|png)/
  or paths starting with known directory prefixes (scripts/, src/,
  config/, web_interface/, systemd/, assets/, docs/, test/, etc.)
- Filtered out URLs, absolute paths (placeholders), and paths
  without slashes (likely not relative refs).
- Checked existence relative to project root.
- Out of 80 unique relative paths in docs/, 32 didn't exist on
  disk. Most were false positives (configkeys mistaken for paths,
  example placeholders like 'assets/myfont.ttf', historical
  references inside docs that already have status banners). The 4
  above were genuine broken refs.

This pattern is reusable for future iterations and worth wiring
into CI (link checker like lychee, scoped to fenced code paths
rather than just markdown links, would catch the same class).

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* docs: address CodeRabbit review comments on #306

Reviewed all 12 CodeRabbit comments on PR #306, verified each against
the current code, and fixed the 11 valid ones. The 12th finding is a
real code bug (cache_manager.delete() calls in api_helper.py and
resource_monitor.py) that's already in the planned follow-up code-fix
PR, so it stays out of this docs PR.

Fixed:

.cursor/plugins_guide.md, .cursor/README.md, .cursorrules
- I claimed "there is no --emulator flag" in 3 places. Verified in
  run.py:19-20 that the -e/--emulator flag is defined and functional
  (it sets os.environ["EMULATOR"]="true" before the display imports).
  Other docs I didn't touch (.cursor/plugin_templates/QUICK_START.md,
  docs/PLUGIN_DEVELOPMENT_GUIDE.md) already use the flag correctly.
  Replaced all 3 wrong statements with accurate guidance that
  both forms work and explains the CLI flag's relationship to the
  env var.

.cursorrules, docs/GETTING_STARTED.md, docs/WEB_INTERFACE_GUIDE.md,
docs/PLUGIN_DEVELOPMENT_GUIDE.md
- Four places claimed "the plugin loader also falls back to plugins/".
  Verified that PluginManager.discover_plugins()
  (src/plugin_system/plugin_manager.py:154) only scans the
  configured directory — no fallback. The fallback to plugins/
  exists only in two narrower places: store_manager.py:1700-1718
  (store install/update/uninstall operations) and
  schema_manager.py:70-80 (schema lookup for the web UI form
  generator). Rewrote all four mentions with the precise scope.
  Added a recommendation to set plugin_system.plugins_directory
  to "plugins" for the smoothest dev workflow with
  dev_plugin_setup.sh symlinks.

docs/FONT_MANAGER.md
- The "Status" warning told plugin authors to use
  display_manager.font_manager.resolve_font(...) as a workaround for
  loading plugin fonts. Verified in src/font_manager.py that
  resolve_font() takes a family name, not a file path — so the
  workaround as written doesn't actually work. Rewrote to tell
  authors to load the font directly with PIL or freetype-py in their
  plugin.
- The same section said "the user-facing font override system in the
  Fonts tab still works for any element that's been registered via
  register_manager_font()". Verified in
  web_interface/blueprints/api_v3.py:5404-5428 that
  /api/v3/fonts/overrides is a placeholder implementation that
  returns empty arrays and contains "would integrate with the actual
  font system" comments — the Fonts tab does not have functional
  integration with register_manager_font() or the override system.
  Removed the false claim and added an explicit note that the tab
  is a placeholder.

docs/ADVANCED_FEATURES.md:523
- The on-demand section said REST/UI calls write a request "into the
  cache manager (display_on_demand_config key)". Wrong — verified
  via grep that api_v3.py:1622 and :1687 write to
  display_on_demand_request, and display_on_demand_config is only
  written by the controller during activation
  (display_controller.py:1195, cleared at :1221). Corrected the key
  name and added controller file:line references so future readers
  can verify.

docs/ADVANCED_FEATURES.md:803
- "Plugins using the background service" paragraph listed all
  scoreboard plugins but an orphaned "⏳ MLB (baseball)" bullet
  remained below from the old version of the section. Removed the
  orphan and added "baseball/MLB" to the inline list for clarity.

web_interface/README.md
- The POST /api/v3/system/action action list was incomplete. Verified
  in web_interface/app.py:1383,1386 that enable_autostart and
  disable_autostart are valid actions. Added both.
- The Plugin Store section was missing
  GET /api/v3/plugins/store/github-status (verified at
  api_v3.py:3296). Added it.
- The SSE line-range reference was app.py:607-615 but line 619
  contains the "Exempt SSE streams from CSRF and add rate limiting"
  block that's semantically part of the same feature. Extended the
  range to 607-619.

docs/GETTING_STARTED.md
- Rows/Columns step said "Columns: 64 or 96 (match your hardware)".
  The web UI's validation accepts any integer in 16-128. Clarified
  that 64 and 96 are the common bundled-hardware values but the
  valid range is wider.

Not addressed (out of scope for docs PR):

- .cursorrules:184 CodeRabbit comment flagged the non-existent
  cache_manager.delete() calls in src/common/api_helper.py:287 and
  src/plugin_system/resource_monitor.py:343. These are real CODE
  bugs, not doc bugs, and they're the first item in the planned
  post-docs-refresh code-cleanup PR (see
  /home/chuck/.claude/plans/warm-imagining-river.md). The docs in
  this PR correctly state that delete() doesn't exist on
  CacheManager — the fix belongs in the follow-up code PR that
  either adds a delete() shim or updates the two callers.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

---------

Co-authored-by: Chuck <chuck@example.com>
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-04-07 21:04:25 -04:00
ChuckandClaude Opus 4.6 77e9eba294 fix: overhaul WiFi captive portal for reliable setup (#296) 
* fix: overhaul WiFi captive portal for reliable device detection and fast setup

The captive portal detection endpoints were returning "success" responses
that told every OS (iOS, Android, Windows, Firefox) that internet was
working — so the portal popup never appeared. This fixes the core issue
and improves the full setup flow:

- Return portal-triggering redirects when AP mode is active; normal
  success responses when not (no false popups on connected devices)
- Add lightweight self-contained setup page (9KB, no frameworks) for
  the captive portal webview instead of the full UI
- Cache AP mode check with 5s TTL (single systemctl call vs full
  WiFiManager instantiation per request)
- Stop disabling AP mode during WiFi scans (which disconnected users);
  serve cached/pre-scanned results instead
- Pre-scan networks before enabling AP mode so captive portal has
  results immediately
- Use dnsmasq.d drop-in config instead of overwriting /etc/dnsmasq.conf
  (preserves Pi-hole and other services)
- Fix manual SSID input bug that incorrectly overwrote dropdown selection

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* fix: address review findings for WiFi captive portal

- Remove orphaned comment left over from old scan_networks() finally block
- Add sudoers rules for dnsmasq drop-in copy/remove to install script
- Combine cached-network message into single showMsg call (was overwriting)
- Return (networks, was_cached) tuple from scan_networks() so API endpoint
  derives cached flag from the scan itself instead of a redundant AP check
- Narrow exception catch in AP mode cache to SubprocessError/OSError and
  log the failure for remote debugging
- Bound checkNewIP retries to 20 attempts (60s) before showing fallback

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-03-27 14:50:33 -04:00
ChuckClaude Sonnet 4.6Chuck
 		fe5c1d0d5e feat(web): add Google Calendar picker widget for dynamic multi-calendar selection (#274) 
* fix(install): add --prefer-binary to pip installs to avoid /tmp exhaustion

timezonefinder (~54 MB) includes large timezone polygon data files that pip
unpacks into /tmp during installation. On Raspberry Pi, the default tmpfs
/tmp size (often ~half of RAM) can be too small, causing the install to fail
with an out-of-space error.

Adding --prefer-binary tells pip to prefer pre-built binary wheels over
source distributions. Since timezonefinder and most other packages publish
wheels on PyPI (and piwheels.org has ARM wheels), this avoids the large
temporary /tmp extraction and speeds up installs generally.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix(timezone): use America/New_York instead of EST for ESPN API date queries

EST is a fixed UTC-5 offset that does not observe daylight saving time,
causing the ESPN API date to be off by one hour during EDT (March–November).
America/New_York correctly handles DST transitions.

The ESPN scoreboard API anchors its schedule calendar to Eastern US time,
so this Eastern timezone is intentionally kept for the API date — it is not
user-configurable. Game time display is converted separately to the user's
configured timezone.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* feat(web): add Google Calendar picker widget for dynamic calendar selection

Adds a new google-calendar-picker widget and API endpoint that lets users
load their available Google Calendars by name and check the ones they want,
instead of manually typing calendar IDs.

- GET /api/v3/plugins/calendar/list-calendars — calls plugin.get_calendars()
  and returns all accessible calendars with id, summary, and primary flag
- google-calendar-picker.js — new widget: "Load My Calendars" button renders
  a checklist; selections update a hidden comma-separated input for form submit
- plugin_config.html — handles x-widget: google-calendar-picker in array branch
- base.html — loads the new widget script

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix(web): address PR review findings in google-calendar-picker

- api_v3.py: replace broad except block with specific exception handling,
  log full traceback via module logger, normalize/validate get_calendars()
  output to stable {id,summary,primary} objects, return opaque user-friendly
  error message instead of leaking str(e)
- google-calendar-picker.js: fix button label only updating to "Refresh
  Calendars" on success (restore original label on error); update summary
  paragraph via syncHiddenAndSummary() on every checkbox change so UI stays
  in sync with hidden input; pass summary element through loadCalendars and
  renderCheckboxes instead of re-querying DOM
- plugin_config.html: bound initWidget retry loop with MAX_RETRIES=40 to
  prevent infinite timers; normalize legacy comma-separated string values
  to arrays before passing to widget.render so pre-existing config populates
  correctly
- install_dependencies_apt.py: update install_via_pip docstring to document
  both --break-system-packages and --prefer-binary flags

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix(web): harden list_calendar_calendars input validation

- Remove unused `as e` binding from ValueError/TypeError/KeyError except clause
- Replace hasattr(__iter__) with isinstance(list|tuple) so non-sequence returns
  are rejected before iteration
- Validate each calendar entry is a collections.abc.Mapping; skip and warn on
  malformed items rather than propagating a TypeError
- Coerce id/summary to str safely if not already strings

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix(web): skip calendar entries with empty id in list_calendar_calendars

After coercing cal_id to str, check it is non-empty before appending to
the calendars list so entries with no usable id are never forwarded to
the client.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

---------

Co-authored-by: Chuck <chuck@example.com>
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
 2026-02-25 18:19:32 -05:00
ChuckChuckClaude Sonnet 4.6
 		8ae82321ce fix(install): add --prefer-binary to pip installs to avoid /tmp exhaustion (#272) 
timezonefinder (~54 MB) includes large timezone polygon data files that pip
unpacks into /tmp during installation. On Raspberry Pi, the default tmpfs
/tmp size (often ~half of RAM) can be too small, causing the install to fail
with an out-of-space error.

Adding --prefer-binary tells pip to prefer pre-built binary wheels over
source distributions. Since timezonefinder and most other packages publish
wheels on PyPI (and piwheels.org has ARM wheels), this avoids the large
temporary /tmp extraction and speeds up installs generally.

Co-authored-by: Chuck <chuck@example.com>
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
 2026-02-25 08:50:35 -05:00
ChuckClaude Sonnet 4.6Chuck
 		23f0176c18 feat: add dev preview server and CLI render script (#264) 
* fix(web): wire up "Check & Update All" plugins button

window.updateAllPlugins was never assigned, so the button always showed
"Bulk update handler unavailable." Wire it to PluginInstallManager.updateAll(),
add per-plugin progress feedback in the button text, show a summary
notification on completion, and skip redundant plugin list reloads.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* feat: add dev preview server, CLI render script, and visual test display manager

Adds local development tools for rapid plugin iteration without deploying to RPi:

- VisualTestDisplayManager: renders real pixels via PIL (same fonts/interface as production)
- Dev preview server (Flask): interactive web UI with plugin picker, auto-generated config
  forms, zoom/grid controls, and mock data support for API-dependent plugins
- CLI render script: render any plugin to PNG for AI-assisted visual feedback loops
- Updated test runner and conftest to auto-detect plugin-repos/ directory

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix(dev-preview): address code review issues

- Use get_logger() from src.logging_config instead of logging.getLogger()
  in visual_display_manager.py to match project logging conventions
- Eliminate duplicate public/private weather draw methods — public draw_sun/
  draw_cloud/draw_rain/draw_snow now delegate to the private _draw_* variants
  so plugins get consistent pixel output in tests vs production
- Default install_deps=False in dev_server.py and render_plugin.py — dev
  scripts don't need to run pip install; developers are expected to have
  plugin deps installed in their venv already
- Guard plugins_dir fixture against PermissionError during directory iteration
- Fix PluginInstallManager.updateAll() to fall back to window.installedPlugins
  when PluginStateManager.installedPlugins is empty (plugins_manager.js
  populates window.installedPlugins independently of PluginStateManager)
- Remove 5 debug console.log statements from plugins_manager.js button setup
  and initialization code

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix(scroll): fix scroll completion to prevent multi-pass wrapping

Change required_total_distance from total_scroll_width + display_width to
total_scroll_width alone. The scrolling image already contains display_width
pixels of blank initial padding, so reaching total_scroll_width means all
content has scrolled off-screen. The extra display_width term was causing
1-2+ unnecessary wrap-arounds, making the same games appear multiple times
and producing a black flicker between passes.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix(dev-preview): address PR #264 code review findings

- docs/DEV_PREVIEW.md: add bash language tag to fenced code block
- scripts/dev_server.py: add MAX/MIN_WIDTH/HEIGHT constants and validate
  width/height in render endpoint; add structured logger calls to
  discover_plugins (missing dirs, hidden entries, missing manifest,
  JSON/OS errors, duplicate ids); add type annotations to all helpers
- scripts/render_plugin.py: add MIN/MAX_DIMENSION validation after
  parse_args; replace prints with get_logger() calls; narrow broad
  Exception catches to ImportError/OSError/ValueError in plugin load
  block; add type annotations to all helpers and main(); rename unused
  module binding to _module
- scripts/run_plugin_tests.py: wrap plugins_path.iterdir() in
  try/except PermissionError with fallback to plugin-repos/
- scripts/templates/dev_preview.html: replace non-focusable div toggles
  with button role="switch" + aria-checked; add keyboard handlers
  (Enter/Space); sync aria-checked in toggleGrid/toggleAutoRefresh
- src/common/scroll_helper.py: early-guard zero total_scroll_width to
  keep scroll_position at 0 and skip completion/wrap logic
- src/plugin_system/testing/visual_display_manager.py: forward color
  arg in draw_cloud -> _draw_cloud; add color param to _draw_cloud;
  restore _scrolling_state in reset(); narrow broad Exception catches in
  _load_fonts to FileNotFoundError/OSError/ImportError; add explicit
  type annotations to draw_text
- test/plugins/test_visual_rendering.py: use context manager for
  Image.open in test_save_snapshot
- test/plugins/conftest.py: add return type hints to all fixtures

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* chore: add bandit and gitleaks pre-commit hooks

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

---------

Co-authored-by: Chuck <chuck@example.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
 2026-02-24 17:57:42 -05:00
ChuckClaude Sonnet 4.5Chuck
 		302235a357 feat: Starlark Apps Integration with Schema-Driven Config + Security Hardening (#253) 
* feat: integrate Starlark/Tronbyte app support into plugin system

Add starlark-apps plugin that renders Tidbyt/Tronbyte .star apps via
Pixlet binary and integrates them into the existing Plugin Manager UI
as virtual plugins. Includes vegas scroll support, Tronbyte repository
browsing, and per-app configuration.

- Extract working starlark plugin code from starlark branch onto fresh main
- Fix plugin conventions (get_logger, VegasDisplayMode, BasePlugin)
- Add 13 starlark API endpoints to api_v3.py (CRUD, browse, install, render)
- Virtual plugin entries (starlark:<app_id>) in installed plugins list
- Starlark-aware toggle and config routing in pages_v3.py
- Tronbyte repository browser section in Plugin Store UI
- Pixlet binary download script (scripts/download_pixlet.sh)

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix(starlark): use bare imports instead of relative imports

Plugin loader uses spec_from_file_location without package context,
so relative imports (.pixlet_renderer) fail. Use bare imports like
all other plugins do.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix(starlark): make API endpoints work standalone in web service

The web service runs as a separate process with display_manager=None,
so plugins aren't instantiated. Refactor starlark API endpoints to
read/write the manifest file directly when the plugin isn't loaded,
enabling full CRUD operations from the web UI.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix(starlark): make config partial work standalone in web service

Read starlark app data from manifest file directly when the plugin
isn't loaded, matching the api_v3.py standalone pattern.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix(starlark): always show editable timing settings in config panel

Render interval and display duration are now always editable in the
starlark app config panel, not just shown as read-only status text.
App-specific settings from schema still appear below when present.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* feat(store): add sort, filter, search, and pagination to Plugin Store and Starlark Apps

Plugin Store:
- Live search with 300ms debounce (replaces Search button)
- Sort dropdown: A→Z, Z→A, Category, Author, Newest
- Installed toggle filter (All / Installed / Not Installed)
- Per-page selector (12/24/48) with pagination controls
- "Installed" badge and "Reinstall" button on already-installed plugins
- Active filter count badge + clear filters button

Starlark Apps:
- Parallel bulk manifest fetching via ThreadPoolExecutor (20 workers)
- Server-side 2-hour cache for all 500+ Tronbyte app manifests
- Auto-loads all apps when section expands (no Browse button)
- Live search, sort (A→Z, Z→A, Category, Author), author dropdown
- Installed toggle filter, per-page selector (24/48/96), pagination
- "Installed" badge on cards, "Reinstall" button variant

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix(store): move storeFilterState to global scope to fix scoping bug

storeFilterState, pluginStoreCache, and related variables were declared
inside an IIFE but referenced by top-level functions, causing
ReferenceError that broke all plugin loading.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* feat(starlark): schema-driven config forms + critical security fixes

## Schema-Driven Config UI
- Render type-appropriate form inputs from schema.json (text, dropdown, toggle, color, datetime, location)
- Pre-populate config.json with schema defaults on install
- Auto-merge schema defaults when loading existing apps (handles schema updates)
- Location fields: 3-part mini-form (lat/lng/timezone) assembles into JSON
- Toggle fields: support both boolean and string "true"/"false" values
- Unsupported field types (oauth2, photo_select) show warning banners
- Fallback to raw key/value inputs for apps without schema

## Critical Security Fixes (P0)
- **Path Traversal**: Verify path safety BEFORE mkdir to prevent TOCTOU
- **Race Conditions**: Add file locking (fcntl) + atomic writes to manifest operations
- **Command Injection**: Validate config keys/values with regex before passing to Pixlet subprocess

## Major Logic Fixes (P1)
- **Config/Manifest Separation**: Store timing keys (render_interval, display_duration) ONLY in manifest
- **Location Validation**: Validate lat [-90,90] and lng [-180,180] ranges, reject malformed JSON
- **Schema Defaults Merge**: Auto-apply new schema defaults to existing app configs on load
- **Config Key Validation**: Enforce alphanumeric+underscore format, prevent prototype pollution

## Files Changed
- web_interface/templates/v3/partials/starlark_config.html — schema-driven form rendering
- plugin-repos/starlark-apps/manager.py — file locking, path safety, config validation, schema merge
- plugin-repos/starlark-apps/pixlet_renderer.py — config value sanitization
- web_interface/blueprints/api_v3.py — timing key separation, safe manifest updates

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

* fix(starlark): use manifest filename field for .star downloads

Tronbyte apps don't always name their .star file to match the directory.
For example, the "analogclock" app has "analog_clock.star" (with underscore).

The manifest.yaml contains a "filename" field with the correct name.

Changes:
- download_star_file() now accepts optional filename parameter
- Install endpoint passes metadata['filename'] to download_star_file()
- Falls back to {app_id}.star if filename not in manifest

Fixes: "Failed to download .star file for analogclock" error

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

* fix(starlark): reload tronbyte_repository module to pick up code changes

The web service caches imported modules in sys.modules. When deploying
code updates, the old cached version was still being used.

Now uses importlib.reload() when module is already loaded.

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

* fix(starlark): use correct 'fileName' field from manifest (camelCase)

The Tronbyte manifest uses 'fileName' (camelCase), not 'filename' (lowercase).
This caused the download to fall back to {app_id}.star which doesn't exist
for apps like analogclock (which has analog_clock.star).

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

* feat(starlark): extract schema during standalone install

The standalone install function (_install_star_file) wasn't extracting
schema from .star files, so apps installed via the web service had no
schema.json and the config panel couldn't render schema-driven forms.

Now uses PixletRenderer to extract schema during standalone install,
same as the plugin does.

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

* feat(starlark): implement source code parser for schema extraction

Pixlet CLI doesn't support schema extraction (--print-schema flag doesn't exist),
so apps were being installed without schemas even when they have them.

Implemented regex-based .star file parser that:
- Extracts get_schema() function from source code
- Parses schema.Schema(version, fields) structure
- Handles variable-referenced dropdown options (e.g., options = dialectOptions)
- Supports Location, Text, Toggle, Dropdown, Color, DateTime fields
- Gracefully handles unsupported fields (OAuth2, LocationBased, etc.)
- Returns formatted JSON matching web UI template expectations

Coverage: 90%+ of Tronbyte apps (static schemas + variable references)

Changes:
- Replace extract_schema() to parse .star files directly instead of using Pixlet CLI
- Add 6 helper methods for parsing schema structure
- Handle nested parentheses and brackets properly
- Resolve variable references for dropdown options

Tested with:
- analog_clock.star (Location field) ✓
- Multi-field test (Text + Dropdown + Toggle) ✓
- Variable-referenced options ✓

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

* fix(starlark): add List to typing imports for schema parser

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

* fix(starlark): load schema from schema.json in standalone mode

The standalone API endpoint was returning schema: null because it didn't
load the schema.json file. Now reads schema from disk when returning
app details via web service.

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

* feat(starlark): implement schema extraction, asset download, and config persistence

## Schema Extraction
- Replace broken `pixlet serve --print-schema` with regex-based source parser
- Extract schema by parsing `get_schema()` function from .star files
- Support all field types: Location, Text, Toggle, Dropdown, Color, DateTime
- Handle variable-referenced dropdown options (e.g., `options = teamOptions`)
- Gracefully handle complex/unsupported field types (OAuth2, PhotoSelect, etc.)
- Extract schema for 90%+ of Tronbyte apps

## Asset Download
- Add `download_app_assets()` to fetch images/, sources/, fonts/ directories
- Download assets in binary mode for proper image/font handling
- Validate all paths to prevent directory traversal attacks
- Copy asset directories during app installation
- Enable apps like AnalogClock that require image assets

## Config Persistence
- Create config.json file during installation with schema defaults
- Update both config.json and manifest when saving configuration
- Load config from config.json (not manifest) for consistency with plugin
- Separate timing keys (render_interval, display_duration) from app config
- Fix standalone web service mode to read/write config.json

## Pixlet Command Fix
- Fix Pixlet CLI invocation: config params are positional, not flags
- Change from `pixlet render file.star -c key=value` to `pixlet render file.star key=value -o output`
- Properly handle JSON config values (e.g., location objects)
- Enable config to be applied during rendering

## Security & Reliability
- Add threading.Lock for cache operations to prevent race conditions
- Reduce ThreadPoolExecutor workers from 20 to 5 for Raspberry Pi
- Add path traversal validation in download_star_file()
- Add YAML error logging in manifest fetching
- Add file size validation (5MB limit) for .star uploads
- Use sanitized app_id consistently in install endpoints
- Use atomic manifest updates to prevent race conditions
- Add missing Optional import for type hints

## Web UI
- Fix standalone mode schema loading in config partial
- Schema-driven config forms now render correctly for all apps
- Location fields show lat/lng/timezone inputs
- Dropdown, toggle, text, color, and datetime fields all supported

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

* fix(starlark): code review fixes - security, robustness, and schema parsing

## Security Fixes
- manager.py: Check _update_manifest_safe return values to prevent silent failures
- manager.py: Improve temp file cleanup in _save_manifest to prevent leaks
- manager.py: Fix uninstall order (manifest → memory → disk) for consistency
- api_v3.py: Add path traversal validation in uninstall endpoint
- api_v3.py: Implement atomic writes for manifest files with temp + rename
- pixlet_renderer.py: Relax config validation to only block dangerous shell metacharacters

## Frontend Robustness
- plugins_manager.js: Add safeLocalStorage wrapper for restricted contexts (private browsing)
- starlark_config.html: Scope querySelector to container to prevent modal conflicts

## Schema Parsing Improvements
- pixlet_renderer.py: Indentation-aware get_schema() extraction (handles nested functions)
- pixlet_renderer.py: Handle quoted defaults with commas (e.g., "New York, NY")
- tronbyte_repository.py: Validate file_name is string before path traversal checks

## Dependencies
- requirements.txt: Update Pillow (10.4.0), PyYAML (6.0.2), requests (2.32.0)

## Documentation
- docs/STARLARK_APPS_GUIDE.md: Comprehensive guide explaining:
  - How Starlark apps work
  - That apps come from Tronbyte (not LEDMatrix)
  - Installation, configuration, troubleshooting
  - Links to upstream projects

All changes improve security, reliability, and user experience.

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

* fix(starlark): convert Path to str in spec_from_file_location calls

The module import helpers were passing Path objects directly to
spec_from_file_location(), which caused spec to be None. This broke
the Starlark app store browser.

- Convert module_path to string in both _get_tronbyte_repository_class
  and _get_pixlet_renderer_class
- Add None checks with clear error messages for debugging

Fixes: spec not found for the module 'tronbyte_repository'

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

* fix(starlark): restore Starlark Apps section in plugins.html

The Starlark Apps UI section was lost during merge conflict resolution
with main branch. Restored from commit 942663ab which had the complete
implementation with filtering, sorting, and pagination.

Fixes: Starlark section not visible on plugin manager page

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

* fix(starlark): restore Starlark JS functionality lost in merge

During the merge with main, all Starlark-specific JavaScript (104 lines)
was removed from plugins_manager.js, including:
- starlarkFilterState and filtering logic
- loadStarlarkApps() function
- Starlark app install/uninstall handlers
- Starlark section collapse/expand logic
- Pagination and sorting for Starlark apps

Restored from commit 942663ab and re-applied safeLocalStorage wrapper
from our code review fixes.

Fixes: Starlark Apps section non-functional in web UI

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

* fix(starlark): security and race condition improvements

Security fixes:
- Add path traversal validation for output_path in download_star_file
- Remove XSS-vulnerable inline onclick handlers, use delegated events
- Add type hints to helper functions for better type safety

Race condition fixes:
- Lock manifest file BEFORE creating temp file in _save_manifest
- Hold exclusive lock for entire read-modify-write cycle in _update_manifest_safe
- Prevent concurrent writers from racing on manifest updates

Other improvements:
- Fix pages_v3.py standalone mode to load config.json from disk
- Improve error handling with proper logging in cleanup blocks
- Add explicit type annotations to Starlark helper functions

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

* fix(starlark): critical bug fixes and code quality improvements

Critical fixes:
- Fix stack overflow in safeLocalStorage (was recursively calling itself)
- Fix duplicate event listeners on Starlark grid (added sentinel check)
- Fix JSON validation to fail fast on malformed data instead of silently passing

Error handling improvements:
- Narrow exception catches to specific types (OSError, json.JSONDecodeError, ValueError)
- Use logger.exception() with exc_info=True for better stack traces
- Replace generic "except Exception" with specific exception types

Logging improvements:
- Add "[Starlark Pixlet]" context tags to pixlet_renderer logs
- Redact sensitive config values from debug logs (API keys, etc.)
- Add file_path context to schema parsing warnings

Documentation:
- Fix markdown lint issues (add language tags to code blocks)
- Fix time unit spacing: "(5min)" -> "(5 min)"

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

* fix(starlark): critical path traversal and exception handling fixes

Path traversal security fixes (CRITICAL):
- Add _validate_starlark_app_path() helper to check for path traversal attacks
- Validate app_id in get_starlark_app(), uninstall_starlark_app(),
  get_starlark_app_config(), and update_starlark_app_config()
- Check for '..' and path separators before any filesystem access
- Verify resolved paths are within _STARLARK_APPS_DIR using Path.relative_to()
- Prevents unauthorized file access via crafted app_id like '../../../etc/passwd'

Exception handling improvements (tronbyte_repository.py):
- Replace broad "except Exception" with specific types
- _make_request: catch requests.Timeout, requests.RequestException, json.JSONDecodeError
- _fetch_raw_file: catch requests.Timeout, requests.RequestException separately
- download_app_assets: narrow to OSError, ValueError
- Add "[Tronbyte Repo]" context prefix to all log messages
- Use exc_info=True for better stack traces

API improvements:
- Narrow exception catches to OSError, json.JSONDecodeError in config loading
- Remove duplicate path traversal checks (now centralized in helper)

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

* fix(starlark): logging improvements and code quality fixes

Logging improvements (pages_v3.py):
- Add logging import and create module logger
- Replace print() calls with logger.warning() with "[Pages V3]" prefix
- Use logger.exception() for outer try/catch with exc_info=True
- Narrow exception handling to OSError, json.JSONDecodeError for file operations

API improvements (api_v3.py):
- Remove unnecessary f-strings (Ruff F541) from ImportError messages
- Narrow upload exception handling to ValueError, OSError, IOError
- Use logger.exception() with context for better debugging
- Remove early return in get_starlark_status() to allow standalone mode fallback
- Sanitize error messages returned to client (don't expose internal details)

Benefits:
- Better log context with consistent prefixes
- More specific exception handling prevents masking unexpected errors
- Standalone/web-service-only mode now works for status endpoint
- Stack traces preserved for debugging without exposing to clients

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

---------

Co-authored-by: Chuck <chuck@example.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
 2026-02-20 19:44:12 -05:00
ChuckClaude Opus 4.6Chuck
 		158e07c82b fix(plugins): prevent root-owned files from blocking plugin updates (#242) 
* fix(web): unify operation history tracking for monorepo plugin operations

The operation history UI was reading from the wrong data source
(operation_queue instead of operation_history), install/update records
lacked version details, toggle operations used a type name that didn't
match UI filters, and the Clear History button was non-functional.

- Switch GET /plugins/operation/history to read from OperationHistory
  audit log with return type hint and targeted exception handling
- Add DELETE /plugins/operation/history endpoint; wire up Clear button
- Add _get_plugin_version helper with specific exception handling
  (FileNotFoundError, PermissionError, json.JSONDecodeError) and
  structured logging with plugin_id/path context
- Record plugin version, branch, and commit details on install/update
- Record install failures in the direct (non-queue) code path
- Replace "toggle" operation type with "enable"/"disable"
- Add normalizeStatus() in JS to map completed→success, error→failed
  so status filter works regardless of server-side convention
- Truncate commit SHAs to 7 chars in details display
- Fix HTML filter options, operation type colors, duplicate JS init

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix(plugins): prevent root-owned files from blocking plugin updates

The root ledmatrix service creates __pycache__ and data cache files
owned by root inside plugin directories. The web service (non-root)
cannot delete these when updating or uninstalling plugins, causing
operations to fail with "Permission denied".

Defense in depth with three layers:
- Prevent: PYTHONDONTWRITEBYTECODE=1 in systemd service + run.py
- Fallback: sudoers rules for rm on plugin directories
- Code: _safe_remove_directory() now uses sudo as last resort,
  and all bare shutil.rmtree() calls routed through it

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix(security): harden sudo removal with path-validated helper script

Address code review findings:

- Replace raw rm/find sudoers wildcards with a vetted helper script
  (safe_plugin_rm.sh) that resolves symlinks and validates the target
  is a strict child of plugin-repos/ or plugins/ before deletion
- Add allow-list validation in sudo_remove_directory() that checks
  resolved paths against allowed bases before invoking sudo
- Check _safe_remove_directory() return value before shutil.move()
  in the manifest ID rename path
- Move stat import to module level in store_manager.py
- Use stat.S_IRWXU instead of 0o777 in chmod fallback stage
- Add ignore_errors=True to temp dir cleanup in finally block
- Use command -v instead of which in configure_web_sudo.sh

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix(security): address code review round 2 — harden paths and error handling

- safe_plugin_rm.sh: use realpath --canonicalize-missing for ALLOWED_BASES
  so the script doesn't fail under set -e when dirs don't exist yet
- safe_plugin_rm.sh: add -- before path in rm -rf to prevent flag injection
- permission_utils.py: use shutil.which('bash') instead of hardcoded /bin/bash
  to match whatever path the sudoers BASH_PATH resolves to
- store_manager.py: check _safe_remove_directory() return before shutil.move()
  in _install_from_monorepo_zip to prevent moving into a non-removed target
- store_manager.py: catch OSError instead of PermissionError in Stage 1 removal
  to handle both EACCES and EPERM error codes
- store_manager.py: hoist sudo_remove_directory import to module level
- configure_web_sudo.sh: harden safe_plugin_rm.sh to root-owned 755 so
  the web user cannot modify the vetted helper script

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix(security): validate command paths in sudoers config and use resolved paths

- configure_web_sudo.sh: validate that required commands (systemctl, bash,
  python3) resolve to non-empty paths before generating sudoers entries;
  abort with clear error if any are missing; skip optional commands
  (reboot, poweroff, journalctl) with a warning instead of emitting
  malformed NOPASSWD lines; validate helper script exists on disk
- permission_utils.py: pass the already-resolved path to the subprocess
  call and use it for the post-removal exists() check, eliminating a
  TOCTOU window between Python-side validation and shell-side execution

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

---------

Co-authored-by: Chuck <chuck@example.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
 2026-02-12 19:28:05 -05:00