specstory-guard

Pass

Audited by Gen Agent Trust Hub on Oct 6, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill performs legitimate security-enhancing tasks. It manages a git pre-commit hook in the local .git/hooks directory to prevent the accidental disclosure of secrets.\n- [COMMAND_EXECUTION]: The skill uses subprocess to execute local git commands (e.g., git rev-parse) and to dispatch internal python scripts (setup.py, scan.py). These operations use argument lists rather than shell strings, which prevents command injection.\n- [DATA_EXPOSURE]: While the skill reads sensitive patterns, it does so to protect the user. It has no network capabilities and cannot exfiltrate the data it scans.\n- [INDIRECT_PROMPT_INJECTION]: The skill scans .specstory/history files which are technically external data. However, the scan results are only used to report findings back to the user/agent and are not executed or evaluated as code, posing a negligible risk.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 6, 2026, 04:22 PM