work
Pass
Audited by Gen Agent Trust Hub on Jun 14, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill orchestrates standard development tools including git, grep, and local application execution for the purpose of code verification and 'oracle' checks. This functionality is essential to its primary purpose as a quality orchestrator.
- [PROMPT_INJECTION]: The skill operates on external codebase data, creating a surface for indirect prompt injection. However, it incorporates robust architectural mitigations: 1. Ingestion points: Reads codebase content via git and the work-sweep sub-skill. 2. Boundary markers: Enforces structured JSON communication for all sub-agent tasks. 3. Capability inventory: Manages sub-agent lifecycles, shell command execution, and file mutations in isolated worktrees. 4. Sanitization: Implements strict verification protocols, notably Lesson L2, which requires agents to independently re-read source files to confirm findings rather than trusting initial evidence.
Audit Metadata