prd-close

Pass

Audited by Gen Agent Trust Hub on Aug 13, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from local PRD files and interpolates it into Forge CLI commands and Git commit messages.
  • Ingestion points: Reads files from the prds/ directory (SKILL.md).
  • Boundary markers: The instructions lack explicit delimiters or warnings to ignore embedded instructions when processing PRD content, though it does confirm the action with the user.
  • Capability inventory: Includes powerful operations such as git mv, git commit to main, git push, and issue updates via gh, glab, or tea (SKILL.md).
  • Sanitization: No explicit sanitization of the PRD file content is performed, however, the shell commands use quoted heredocs (cat <<'EOF') which prevents the shell from interpreting content as commands or variables during interpolation.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 13, 2026, 05:40 PM
Security Audit — agent-trust-hub — prd-close