ClawAudit verdict
pao-system
The skill describes a Python-based personal AI operator framework with WebSocket communication; the capabilities described (skill management, protocol handling) are consistent with the stated purpose and no exfiltration or malicious behavior is evident.
⚠ 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 (11)
Opens WebSocket connection
SKILL.md · frontmatter · WebSocket
subprocess execution — runs system commands from Python
deployment/installer.py · prose · downgraded · subprocess.run(
Uses exec() — may execute shell commands
htmlcov/coverage_html.js · prose · downgraded · exec(
Python httpx request — network access
README.md · code · httpx.post(
POSTs data to external URL
README.md · code · .post('http://
Python os.environ.get — reads environment variable
build/lib/core/config.py · prose · downgraded · os.environ.get(
Python shutil file operation — copies/moves/deletes files
deployment/installer.py · prose · downgraded · shutil.rmtree(
References webhook/callback URL
src/protocols/task_protocol.py · prose · downgraded · callback_url
References tunneling service
task_deploy_guide.md · prose · downgraded · ngrok
Python aiohttp session — async network access
test_all.py · prose · downgraded · aiohttp.ClientSession
Python os.getenv — reads environment variable
test_bidirectional.py · prose · downgraded · os.getenv(
Why the tier is capped
Execution sink present in raw bytes (Hard Floor: class D). 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.