n8n-monitor-hardened
Pass
Audited by Gen Agent Trust Hub on Apr 21, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes read-only Docker CLI commands (
ps,logs,inspect,stats) for system monitoring. Execution is strictly scoped to containers matching the 'n8n' string to prevent unauthorized access to adjacent infrastructure components. - [DATA_EXFILTRATION]: A documentation file,
SAFETY.md, contains a reference to a network exfiltration vector (/dev/tcp/10.0.0.5/4444). This string is explicitly identified as a 'FAIL' case to instruct the agent to refuse similar commands and ensure all data remains within the conversation context. - [PROMPT_INJECTION]: The skill processes external data from Docker logs, which represents a potential indirect injection surface. This risk is mitigated by comprehensive instructions in
SKILL.mdthat mandate the sanitization of output and the redaction of secrets, such as API keys and credentials, regardless of user authority or context.
Audit Metadata