hermes-agent

Pass

Audited by Gen Agent Trust Hub on Apr 12, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The scripts/install_hermes.sh script fetches the Hermes Agent source code from the official NousResearch GitHub repository (https://github.com/NousResearch/hermes-agent.git). This is a reputable source consistent with the skill's stated purpose.
  • [COMMAND_EXECUTION]: Shell scripts (hermes_wrapper.sh, hermes_delegate.sh) wrap the hermes CLI for integration with AI agents. These scripts execute the local binary and manage process lifecycle (timeouts, locks) using standard bash patterns.
  • [SAFE]: Static analysis flagged destructive command strings in references/config-guide.md. Technical review confirms these strings are part of a denied_commands denylist (e.g., rm -rf /, sudo, chmod 777) intended to enhance the security of the agent's terminal tool.
  • [SAFE]: The skill implements a local telemetry mechanism in scripts/hermes_delegate.sh that logs execution errors to ~/.workbuddy/memory/self-improving/learnings.jsonl. This data is stored locally for the purpose of 'self-improving' agent logic and shows no signs of external exfiltration.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 12, 2026, 06:25 AM