brief
Pass
Audited by Gen Agent Trust Hub on Nov 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input to generate structured briefs for autonomous agents.
- Ingestion points: Verbatim user intent captured in Step 1 of SKILL.md.
- Boundary markers: Employs markdown headers and HTML comments to separate fields, though the core intent is unescaped.
- Capability inventory: Includes filesystem writes, GitHub API calls (gh api), and command delegation to runners.
- Sanitization: No explicit filtering or sanitization of user-provided content is performed.
- [COMMAND_EXECUTION]: Utilizes GitHub CLI (gh) commands to manage issue tracking and sub-task coordination. This includes API calls for linking issues and executing runner commands like /lever deliver.
Audit Metadata