ClawAudit verdict
adversarial-engine
openclaw-adversarial-engine
The skill uses network_in capability and executionSinkDetected is true, but the content seems to describe a legitimate use case for an adversarial engine.
⚠ 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 (4)
Possible hardcoded credential
async_engine.py · prose · downgraded · API_KEY = "sk-sp-2b89d1b9a55d4cb9a8094c9127459aab
Opens WebSocket connection
SKILL.md · code · WebSocket
subprocess execution — runs system commands from Python
async_engine.py · prose · downgraded · subprocess.run(
Python aiohttp session — async network access
async_engine.py · prose · downgraded · aiohttp.ClientSession
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.
network_in Is this flag fair?
Thanks — recorded.