myclaw-guardian

Pass

Audited by Gen Agent Trust Hub on Mar 7, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill deploys a bash script (guardian.sh) that manages processes and file versions. It uses pgrep and pkill to monitor and restart the openclaw-gateway process. It also utilizes git commands (reset --hard, commit, add) within the configured workspace to provide automated rollback and backup functionality.
  • [EXTERNAL_DOWNLOADS]: The script contains logic to send POST requests via curl to a Discord webhook URL provided by the user through the DISCORD_WEBHOOK_URL environment variable. This is used solely for status notifications and does not involve downloading executable code.
  • [SAFE]: No malicious patterns, obfuscation, or unauthorized data access were detected. The skill's operations are transparent and align with its documented purpose of providing a resilience layer for the OpenClaw Gateway.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 7, 2026, 09:32 PM