fetch-pr-feedback

Pass

Audited by Gen Agent Trust Hub on May 18, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill uses official GitHub CLI tools (gh) and standard JSON processing utilities (jq) to fetch and filter PR metadata and comments. No unauthorized file access, credential harvesting, or suspicious network exfiltration was detected.
  • [PROMPT_INJECTION]: The skill exhibits a potential surface for indirect prompt injection because it ingests PR comments which are provided by external users. This is an inherent risk factor for the skill's primary purpose.
  • Ingestion points: External comments are fetched via gh api in SKILL.md (Step 3).
  • Boundary markers: The comments are formatted into a markdown document but are not enclosed in specific boundary markers or safety instructions to prevent the model from following commands embedded in the comments.
  • Capability inventory: The skill uses the GitHub CLI to read PR data and orchestrates the receive-feedback skill for further evaluation.
  • Sanitization: The implementation includes a clean_body function using jq to strip bot-specific HTML comments and noise, although it does not filter the natural language content for adversarial instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
May 18, 2026, 04:42 AM
Security Audit — agent-trust-hub — fetch-pr-feedback