skills/jhostalek/dotclaude/commit/Gen Agent Trust Hub

commit

Pass

Audited by Gen Agent Trust Hub on Jun 16, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes standard Git commands (git diff, git add, git commit, git reset) to manage repository state. These operations are appropriate for the skill's stated purpose of managing commits.
  • [DYNAMIC_CONTEXT_INJECTION]: The instruction !cat ~/.claude/skills/conventional-commits.md`` is used to dynamically load documentation into the skill's context at load time. This is used here to import conventional commit guidelines from a local configuration directory.
  • [INDIRECT_PROMPT_INJECTION]: By reading the output of git diff, the skill identifies a surface where untrusted data from the repository could potentially influence agent behavior. This is documented as a standard risk factor for tools that process code deltas.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 16, 2026, 04:52 AM
Security Audit — agent-trust-hub — commit