pr-triage

Pass

Audited by Gen Agent Trust Hub on Aug 10, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses git, GitHub CLI (gh), and GitLab CLI (glab) to interact with code repositories and forge platforms. This behavior is necessary for the skill's stated purpose of PR triage.\n- [EXTERNAL_DOWNLOADS]: Fetches pull request branches from GitHub and GitLab. These operations target well-known services and are standard for development workflows.\n- [PROMPT_INJECTION]: The skill processes untrusted external data (PR diffs, descriptions, and Discord messages), creating a surface for indirect prompt injection.\n
  • Ingestion points: PR descriptions, diffs, comments, and Discord threads (SKILL.md).\n
  • Boundary markers: Mentions the use of >>>...<<< for Discord threads.\n
  • Capability inventory: Subprocess calls to git, gh, and glab (SKILL.md, reference.md).\n
  • Sanitization: Relies on an explicit user 'go' command before any write actions are performed.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 10, 2026, 11:08 PM
Security Audit — agent-trust-hub — pr-triage