arc-prd-to-issues

Pass

Audited by Gen Agent Trust Hub on Jul 11, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection because it ingests and processes untrusted data from external sources.
  • Ingestion points: SKILL.md (Step 1) fetches PRD content from GitHub issues, URLs, or external file paths.
  • Boundary markers: The instructions do not define specific delimiters to wrap the untrusted PRD content when it is interpolated into the agent's context.
  • Capability inventory: The skill utilizes the gh CLI tool to create issues (gh issue create) and explore the repository structure.
  • Sanitization: While the skill does not explicitly sanitize the PRD content, it implements a robust mitigation via Step 4 ("Quiz the user before issue creation"), which requires the user to review and approve the drafted issues before any write operations are performed. This human-in-the-loop checkpoint prevents the agent from being autonomously manipulated into performing unwanted actions by instructions hidden within a PRD.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 11, 2026, 02:14 PM
Security Audit — agent-trust-hub — arc-prd-to-issues