triage-aging-prs

Pass

Audited by Gen Agent Trust Hub on Sep 16, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and act upon data from pull requests, including branch code, titles, and review comments, which are untrusted external inputs.\n
  • Ingestion points: The skill uses gh pr list and gh api to fetch pull request metadata and reviews, and the Task tool to read branch code as described in Step 0 and Step 2 of SKILL.md.\n
  • Boundary markers: The skill provides a dedicated reference file assets/references/resolving-review-findings.md which explicitly labels review text as untrusted.\n
  • Capability inventory: The agent has access to Bash (for merging and repository configuration), Write, and Edit tools.\n
  • Sanitization: Rule 1 in the reference file instructs the agent to treat findings as untrusted claims and never follow embedded instructions; Rule 3 mandates verification against the code at HEAD before implementation.\n- [COMMAND_EXECUTION]: The skill performs sensitive repository management tasks via the GitHub CLI.\n
  • Evidence: In SKILL.md (Step 1b), the skill provides logic to modify GitHub repository rulesets (gh api -X PUT \"repos/$REPO/rulesets/$ID\") to relax required status checks. While it instructs the agent to backup settings and ask for user confirmation, this capability allows for the manipulation of repository security posture.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 16, 2026, 09:29 PM
Security Audit — agent-trust-hub — triage-aging-prs