git-guardrails

Pass

Audited by Gen Agent Trust Hub on Nov 8, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses subprocess.run in assets/git_guardrails_hook.py to call git symbolic-ref for the purpose of identifying the current branch. This is a read-only operation and essential for determining if a branch is protected.\n- [COMMAND_EXECUTION]: The installer script scripts/install_guardrails.py and the tester scripts/guardrail_test.py use subprocess.run to execute local scripts and git commands as part of their installation and verification workflows. These are routine operations for a developer tool.\n- [PERSISTENCE]: The skill's primary function is to install hooks into .claude/settings.json and .git/hooks/. While this involves modifying configuration files to ensure the guardrail persists across sessions, the behavior is transparently documented and serves its stated safety purpose.\n- [PRIVILEGE_ESCALATION]: The installer uses chmod to ensure that the installed hook scripts are executable. This is standard practice for script installation and is limited to the skill's own assets.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input (the command string the agent intends to run) through the git_guardrails_hook.py script. It uses shlex for safe tokenization and parsing to avoid common shell injection pitfalls. It follows a 'fail-open' design for unparseable input to ensure user workflow is not unnecessarily disrupted.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 8, 2026, 03:14 PM