arc-creating-user-stories

Pass

Audited by Gen Agent Trust Hub on Jul 12, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes the official gh CLI to automate tasks like creating issues and labels. It specifically mitigates command injection risks by instructing the agent to write markdown content to temporary files and use the --body-file flag instead of passing raw content via inline arguments.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests repository files and specifications which serves as an attack surface for indirect prompt injection.
  • Ingestion points: Processes repository files, PRDs, specs, and existing issues in SKILL.md (Step 1).
  • Boundary markers: Includes a mandatory user review step in SKILL.md (Step 3) to verify the drafted backlog before any GitHub resources are created.
  • Capability inventory: Features the ability to create labels, issues, and projects using the gh tool as defined in GITHUB.md.
  • Sanitization: Employs file-based communication for CLI commands to ensure external content is not interpreted as shell commands.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 12, 2026, 05:30 PM
Security Audit — agent-trust-hub — arc-creating-user-stories