draft-pr

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill is configured to fetch instruction metadata from the vendor's repository (spacemansh/spaceman-skills) when local sibling skills are unavailable.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted local data including code diffs and repository-specific PR templates, which represents a potential injection surface.
  • Ingestion points: Reads repository diffs and pull request templates from .github/ (SKILL.md Workflow Steps 1 and 5).
  • Boundary markers: No explicit delimiter usage is specified for separating instructions from processed data.
  • Capability inventory: Perform git pushes and GitHub CLI operations (git push, gh pr create, gh pr edit).
  • Sanitization: Includes a preflight check to detect and stop execution if credentials, tokens, or .env values are identified within the diff.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 05:51 PM