ClawAudit verdict
kai-report-creator
Generates self-contained HTML reports from user-provided data entirely locally; no network calls, no credential access, and no suspicious execution patterns.
⚠ Flagged for review — coarse, uncorroborated signal, not a confirmed exploit. Review the config yourself before installing.
Automated static analysis — not a human review. ClawAudit flags capabilities, not confirmed intent, and can produce false positives. Disagree with this verdict? Use Dispute below.
Findings (7)
<script> tag in markdown — potential code injection
examples/zh/kai-report-creator-guide.report.md · code · <script
Writes to SKILL.md — self-modifying skill
README.md · prose · downgraded · SKILL.md pre-write
subprocess execution — runs system commands from Python
scripts/run-skill-evals.py · prose · downgraded · subprocess.run(
Instructs covert action — may act without user awareness
tests/test_html_shell_contract.py · prose · downgraded · silently
References agent configuration files
CLAUDE.md · prose · downgraded · CLAUDE.md
Python shutil file operation — copies/moves/deletes files
scripts/clean-generated.py · prose · downgraded · shutil.rmtree(
Popular HTTP library — network access
tests/test_theme_screenshot_assets.py · prose · downgraded · got
Why the tier is capped
Execution sink present in raw bytes (Hard Floor: class B/D/G). Final tier capped at Caution — cannot be lifted by any downgrade, example-payload opt-in, or allowlist.
Permissions & capabilities
No declared permissions — minimal attack surface.
Is this flag fair?
Thanks — recorded.