git-guardrails-claude-code
Pass
Audited by Gen Agent Trust Hub on Apr 15, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill sets up a local interceptor script (
block-dangerous-git.sh) that evaluates commands before they are executed by the agent. This is used here as a security guardrail to prevent accidental data loss or destructive git operations.\n- [SAFE]: The skill modifies environment-specific configuration files (.claude/settings.json) to register its hook. This is a standard and documented method for extending the agent's workflow safety.\n- [SAFE]: The distributed bash script uses standard tools likejqandgrepto perform pattern matching on commands and does not contain any malicious logic, network calls, or hidden behavior.
Audit Metadata