Skills
Navigate
skills/o-marsters-1997/my-claude-code/git-guardrails-claude-code/Gen Agent Trust Hub

git-guardrails-claude-code

Pass

Audited by Gen Agent Trust Hub on Jun 14, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill installs a local bash script (block-dangerous-git.sh) and configures it as a Claude Code hook. This involves making the script executable via chmod +x and modifying the .claude/settings.json configuration file to intercept bash commands.
  • [SAFE]: The primary function of the skill is a defensive security guardrail designed to prevent accidental data loss or destructive operations. The script operates entirely on the local machine and does not perform network operations or access sensitive credentials.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 14, 2026, 07:06 PM