monitor-claude

Pass

Audited by Gen Agent Trust Hub on Mar 17, 2026

Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses ps aux to gather system-wide process data and os.kill to terminate processes that exceed resource thresholds. It also interacts with system services using systemctl and busctl to verify the health of the Embry Agent daemon.
  • [DATA_EXFILTRATION]: Transmits process metadata, including PIDs, memory usage (RSS), CPU percentages, and truncated command strings, to an external Discord webhook for remote health monitoring.
  • [EXTERNAL_DOWNLOADS]: Includes httpx in its dependencies to facilitate network communication with the Discord API for sending alerts.
  • [SAFE]: All operations are consistent with the skill's documented purpose of process management and health monitoring. No obfuscation, privilege escalation, or unauthorized data access patterns were detected.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 17, 2026, 06:37 AM
Security Audit — agent-trust-hub — monitor-claude