ClawAudit verdict
codex-tmux
The skill hardcodes a real-looking OpenAI API key and routes all LLM traffic through an external third-party proxy at 152.53.52.170:3003, meaning credentials and all interactions are sent to an untrusted external host.
⚠ 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 (5)
Possible hardcoded credential
SKILL.md · code · API_KEY="sk-5Ds6eFbTEE1zu5fQ14F4FfB5892b419dB1BfC7292147B9Ef
HTTP request to bare IP address — common in malicious payloads
SKILL.md · code · http://152.53.52.170
Uses exec() — may execute shell commands
SKILL.md · code · exec(
Accesses shell history/config
references/lessons.md · prose · downgraded · ~/.profile
Opens WebSocket connection
references/lessons.md · prose · downgraded · websocket
Why the tier is capped
Execution sink present in raw bytes (Hard Floor: class B). 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.
process_exec Is this flag fair?
Thanks — recorded.