ClawAudit verdict
checktruth
Fact-checking skill using only the host LLM and web search for core functionality; clearly documents that optional external LLM API keys are not required and not loaded by default, with explicit data handling transparency.
⚠ 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 (3)
Possible hardcoded credential
reference/README.md · code · API_KEY="your-glm-api-key
Python os.getenv — reads environment variable
reference/multi_model_verify.py · prose · downgraded · os.getenv(
POSTs data to external URL
reference/src/fetch_answers.py · prose · downgraded · .post(
"https://
Permissions & capabilities
No declared permissions — minimal attack surface.
credential_access Is this flag fair?
Thanks — recorded.