skills/mrwogu/promptscript/committing/Gen Agent Trust Hub

committing

Pass

Audited by Gen Agent Trust Hub on Mar 5, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The instructions are strictly focused on git workflow and do not contain any attempts to override system behavior, bypass filters, or ignore previous instructions.
  • [CREDENTIALS_UNSAFE]: No hardcoded credentials or secrets were found. The skill explicitly includes a safety section advising users never to commit secrets, credentials, or environment files.
  • [COMMAND_EXECUTION]: The skill uses standard git commands (status, diff, log, commit). It correctly implements a quoted HEREDOC pattern (cat <<'EOF') for the commit command, which is a security best practice to prevent shell expansion or command injection from the commit message content.
  • [EXTERNAL_DOWNLOADS]: No remote scripts, external packages, or unauthorized downloads are requested or executed.
  • [DATA_EXFILTRATION]: There are no network operations or indicators of data being sent to external domains.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 5, 2026, 08:50 PM