openclaw-debugging
Pass
Audited by Gen Agent Trust Hub on May 14, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
- [COMMAND_EXECUTION]: The instructions direct the agent to run project-specific commands like
openclaw gatewayandpnpm buildto verify system behavior during debugging.\n- [DATA_EXFILTRATION]: The skill mentions checking for existing API keys in the user's~/.profilefile but explicitly commands the agent to never print secrets, mitigating the risk of credential exposure.
Audit Metadata