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>
This commit is contained in:
Chuck
2026-07-15 14:33:17 -04:00
committed by GitHub
co-authored by Claude Sonnet 5
parent 4abcd0e4f9
commit 66f9950a30
4 changed files with 1322 additions and 0 deletions
+344
View File
@@ -0,0 +1,344 @@
#!/usr/bin/env python3
"""
LEDMatrix Plugin Security Auditor
Performs AST-based security analysis of all Python files in plugin directories.
Designed to run in CI — exits non-zero on CRITICAL findings only.
Usage:
python scripts/audit_plugins.py
python scripts/audit_plugins.py --verbose
python scripts/audit_plugins.py --plugin hello-world
python scripts/audit_plugins.py --output results.json
"""
import ast
import argparse
import json
import sys
from dataclasses import dataclass, asdict
from pathlib import Path
from datetime import datetime, timezone
PROJECT_ROOT = Path(__file__).resolve().parent.parent
PLUGIN_BASE_DIRS = [
PROJECT_ROOT / "plugins",
PROJECT_ROOT / "plugin-repos",
]
# ─────────────────────────────────────────────────────────────────────────────
# Finding dataclass
# ─────────────────────────────────────────────────────────────────────────────
@dataclass
class Finding:
plugin_id: str
file: str
line: int
severity: str # CRITICAL | WARNING | INFO
rule: str
message: str
def to_dict(self) -> dict:
return asdict(self)
# ─────────────────────────────────────────────────────────────────────────────
# AST visitor
# ─────────────────────────────────────────────────────────────────────────────
class _PluginVisitor(ast.NodeVisitor):
"""Collect security findings from a single plugin Python file."""
def __init__(self, filepath: Path, plugin_id: str):
self.filepath = filepath
self.plugin_id = plugin_id
self.findings: list[Finding] = []
# Local name -> real dotted path, so aliased imports and from-imports
# of dangerous APIs (import subprocess as sp; from builtins import
# eval as e) are still recognized in visit_Call below.
self._aliases: dict[str, str] = {}
def _add(self, node: ast.AST, severity: str, rule: str, message: str) -> None:
self.findings.append(Finding(
plugin_id=self.plugin_id,
file=str(self.filepath.relative_to(PROJECT_ROOT)),
line=getattr(node, "lineno", 0),
severity=severity,
rule=rule,
message=message,
))
def _resolve(self, local_name: str) -> str:
"""Resolve a local name through recorded import aliases to its real
dotted path (e.g. "sp" -> "subprocess"); unresolved names pass through
unchanged."""
return self._aliases.get(local_name, local_name)
def _resolve_call_target(self, func: ast.expr) -> str | None:
"""Resolve a Call's func node to a fully-qualified dotted target,
covering a direct name (bare builtin, aliased import, or
from-import: from builtins import eval as e; from subprocess
import run; from os import system as s) and module-attribute
access (subprocess.run, sp.run, os.system, o.system) uniformly.
Returns None for call shapes this doesn't attempt to resolve."""
if isinstance(func, ast.Name):
return self._resolve(func.id)
if isinstance(func, ast.Attribute) and isinstance(func.value, ast.Name):
base = self._resolve(func.value.id)
return f"{base}.{func.attr}"
return None
def visit_Call(self, node: ast.Call) -> None:
target = self._resolve_call_target(node.func)
if target is None:
self.generic_visit(node)
return
leaf = target.rsplit(".", 1)[-1]
# eval() / exec() / compile() — arbitrary code execution, whether a
# bare call, an aliased import, or a from-import
# (from builtins import eval as e; e(...))
if leaf == "eval":
self._add(node, "CRITICAL", "PLUGIN-001",
"eval() call — arbitrary code execution risk")
elif leaf == "exec":
self._add(node, "CRITICAL", "PLUGIN-002",
"exec() call — arbitrary code execution risk")
elif leaf == "compile":
self._add(node, "WARNING", "PLUGIN-003",
"compile() call — dynamic code compilation")
# subprocess.*(shell=True), whether subprocess.run(...), sp.run(...),
# or a from-import (from subprocess import run; run(..., shell=True))
if target in {
"subprocess.run", "subprocess.call", "subprocess.Popen",
"subprocess.check_call", "subprocess.check_output",
}:
for kw in node.keywords:
if (kw.arg == "shell" and
isinstance(kw.value, ast.Constant) and
kw.value.value is True):
self._add(node, "WARNING", "PLUGIN-004",
f"subprocess.{leaf}(shell=True) — "
f"shell injection risk if args include user input")
# os.system(), whether os.system(...), o.system(...), or a
# from-import (from os import system as s; s(...))
if target == "os.system":
self._add(node, "WARNING", "PLUGIN-005",
"os.system() call — prefer subprocess with list args")
self.generic_visit(node)
def visit_Import(self, node: ast.Import) -> None:
for alias in node.names:
if alias.asname:
local, real = alias.asname, alias.name
else:
# `import os.path` binds the top-level name `os`, not `os.path`
local = real = alias.name.split(".")[0]
self._aliases[local] = real
self._check_import(node, alias.name)
self.generic_visit(node)
def visit_ImportFrom(self, node: ast.ImportFrom) -> None:
if node.module:
for alias in node.names:
local = alias.asname or alias.name
self._aliases[local] = f"{node.module}.{alias.name}"
self._check_import(node, node.module)
self.generic_visit(node)
def _check_import(self, node: ast.AST, module_name: str) -> None:
dangerous = {
"ctypes": ("WARNING", "PLUGIN-010", "ctypes import — native code execution"),
"cffi": ("WARNING", "PLUGIN-011", "cffi import — native code execution"),
"pickle": ("WARNING", "PLUGIN-012",
"pickle import — deserialization can execute arbitrary code"),
"marshal": ("WARNING", "PLUGIN-013",
"marshal import — deserialization risk"),
}
for mod, (severity, rule, msg) in dangerous.items():
if module_name == mod or module_name.startswith(mod + "."):
self._add(node, severity, rule, msg)
# ─────────────────────────────────────────────────────────────────────────────
# Per-plugin audit
# ─────────────────────────────────────────────────────────────────────────────
def audit_plugin(plugin_dir: Path) -> list[Finding]:
"""Audit a single plugin directory. Returns all findings."""
findings: list[Finding] = []
plugin_id = plugin_dir.name
# Check for required files
for required_file, rule, msg in [
("manifest.json", "PLUGIN-020",
"manifest.json missing — plugin may be incomplete"),
("config_schema.json", "PLUGIN-021",
"config_schema.json missing — no input validation schema declared"),
]:
if not (plugin_dir / required_file).exists():
findings.append(Finding(
plugin_id=plugin_id,
file=str((plugin_dir / required_file).relative_to(PROJECT_ROOT)),
line=0,
severity="WARNING",
rule=rule,
message=msg,
))
# AST analysis of all Python files
for py_file in sorted(plugin_dir.rglob("*.py")):
try:
source = py_file.read_text(encoding="utf-8")
tree = ast.parse(source, filename=str(py_file))
visitor = _PluginVisitor(py_file, plugin_id)
visitor.visit(tree)
findings.extend(visitor.findings)
except SyntaxError as exc:
# A file the visitor can't even parse is a file we can't verify
# is safe -- this must block the audit, not just warn.
findings.append(Finding(
plugin_id=plugin_id,
file=str(py_file.relative_to(PROJECT_ROOT)),
line=getattr(exc, "lineno", 0) or 0,
severity="CRITICAL",
rule="PLUGIN-030",
message=f"Python syntax error — cannot be parsed: {exc}",
))
except OSError as exc:
# Same reasoning as SyntaxError: an unreadable file was never
# actually scanned, so it must block rather than pass silently.
findings.append(Finding(
plugin_id=plugin_id,
file=str(py_file.relative_to(PROJECT_ROOT)),
line=0,
severity="CRITICAL",
rule="PLUGIN-031",
message=f"Could not read file: {exc}",
))
return findings
# ─────────────────────────────────────────────────────────────────────────────
# Main
# ─────────────────────────────────────────────────────────────────────────────
def main() -> int:
parser = argparse.ArgumentParser(
description="LEDMatrix plugin security auditor",
formatter_class=argparse.RawDescriptionHelpFormatter,
)
parser.add_argument("--plugin", "-p", default=None,
help="Audit a specific plugin ID only")
parser.add_argument("--output", "-o", default=None,
help="Write JSON results to this file")
parser.add_argument("--verbose", "-v", action="store_true",
help="Show all findings, not just summary")
args = parser.parse_args()
print("=" * 60)
print("LEDMatrix Plugin Security Audit")
print(f"Project root: {PROJECT_ROOT}")
print("=" * 60)
all_findings: list[Finding] = []
plugins_scanned = 0
plugin_found = args.plugin is None
for base_dir in PLUGIN_BASE_DIRS:
if not base_dir.exists():
if args.verbose:
print(f" ⏭️ Skipping {base_dir.name}/ (directory not found)")
continue
base_label = base_dir.relative_to(PROJECT_ROOT)
print(f"\n Scanning {base_label}/")
for plugin_dir in sorted(base_dir.iterdir()):
if not plugin_dir.is_dir():
continue
if plugin_dir.name.startswith((".", "_")):
continue
if args.plugin and plugin_dir.name != args.plugin:
continue
if args.plugin:
plugin_found = True
findings = audit_plugin(plugin_dir)
all_findings.extend(findings)
plugins_scanned += 1
critical = [f for f in findings if f.severity == "CRITICAL"]
warnings = [f for f in findings if f.severity == "WARNING"]
if critical:
icon, label = "🚨", "CRITICAL"
elif warnings:
icon, label = "⚠️ ", "WARN "
else:
icon, label = "", "PASS "
print(f" {icon} [{label}] {plugin_dir.name}"
f"{len(critical)} critical, {len(warnings)} warnings")
if args.verbose:
for f in findings:
severity_icon = {"CRITICAL": "🚨", "WARNING": "⚠️ ", "INFO": ""}.get(
f.severity, " "
)
print(f" {severity_icon} {f.rule} {f.file}:{f.line}{f.message}")
if args.plugin and not plugin_found:
print(f"\n 🚨 Plugin '{args.plugin}' not found in any of "
f"{[str(d.relative_to(PROJECT_ROOT)) for d in PLUGIN_BASE_DIRS]}"
f"nothing was audited")
return 1
# Summary
critical_findings = [f for f in all_findings if f.severity == "CRITICAL"]
warning_findings = [f for f in all_findings if f.severity == "WARNING"]
print(f"\n{'=' * 60}")
print(f" Plugins scanned : {plugins_scanned}")
print(f" CRITICAL : {len(critical_findings)}")
print(f" WARNING : {len(warning_findings)}")
if critical_findings:
print("\n 🚨 CRITICAL findings:")
for f in critical_findings:
print(f" {f.plugin_id} | {Path(f.file).name}:{f.line} | {f.message}")
# Write JSON output
if args.output:
output_data = {
"timestamp": datetime.now(timezone.utc).isoformat(),
"plugins_scanned": plugins_scanned,
"summary": {
"critical": len(critical_findings),
"warnings": len(warning_findings),
},
"findings": [f.to_dict() for f in all_findings],
}
Path(args.output).write_text(
json.dumps(output_data, indent=2), encoding="utf-8"
)
print(f"\n Results written to: {args.output}")
if critical_findings:
print("\n 🚨 Blocking — CRITICAL issues must be resolved")
return 1
print("\n ✅ No critical issues found")
return 0
if __name__ == "__main__":
sys.exit(main())