ClawAudit verdict
lobster-market
AI agent marketplace skill for registering and invoking agents on Lobster Market; credential_access reads JWT tokens and API keys stored locally in ~/.lobster-market/, all operations are transparent and user-directed through the declared mindcore8.com gateway.
⚠ 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.
Permission integrity
credential_access
Findings (8)
Pipe to python — executes piped content as Python code
scripts/start-agents.sh · prose · downgraded · | python3
Pipe-to-python pattern — remote code execution risk
scripts/start-agents.sh · prose · downgraded · curl -sf "http://localhost:$port/health" 2>/dev/null | python
Python os.environ.get — reads environment variable
scripts/adapters/llm_adapter.py · prose · downgraded · os.environ.get(
Python urllib.request — network access
scripts/adapters/llm_adapter.py · prose · downgraded · urllib.request
pip3 install — installs Python packages at runtime
scripts/adapters/nanobot_adapter.py · prose · downgraded · pip3 install
Python http.client — stdlib network access
scripts/lobster.py · prose · downgraded · http.client.HTTPSConnection
Opens WebSocket connection
scripts/market-connect.py · prose · downgraded · WebSocket
Python aiohttp session — async network access
scripts/market-connect.py · prose · downgraded · aiohttp.ClientSession
Why the tier is capped
Execution sink present in raw bytes (Hard Floor: class C/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.
credential_accessnetwork_in Is this flag fair?
Thanks — recorded.