git-guardrails-claude-code
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is designed to implement protective measures. It identifies patterns of destructive Git commands (e.g.,
git push --force,git reset --hard) and prevents their execution through a local hook mechanism.\n- [COMMAND_EXECUTION]: The skill facilitates the setup of a local bash script (block-dangerous-git.sh) and configures thePreToolUsehook in Claude's settings. This is a legitimate use of the platform's extensibility features to manage command permissions.\n- [PRIVILEGE_ESCALATION]: The process involves making the local hook script executable usingchmod +x. This is a standard administrative step for the intended functionality and does not involve acquiring unauthorized system privileges.
Audit Metadata