acceptance-criteria-verification
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from GitHub issue bodies (SKILL.md Step 1) and uses it to plan verification tasks and update issue content. Without boundary markers or sanitization, instructions embedded in issues could influence the agent's behavior. \n
- Ingestion points: Issue body fetched via
gh issue view. \n - Boundary markers: None specified. \n
- Capability inventory: Shell execution (gh, git, pnpm, npx) and browser automation. \n
- Sanitization: None present.\n- [COMMAND_EXECUTION]: The skill performs shell operations like running tests and managing GitHub issues. These commands use external data as input, which can be risky if not carefully handled by the agent.
Audit Metadata