review
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted test code files to perform quality reviews, which creates a potential surface for instructions embedded in test files to influence agent behavior.
- Ingestion points: Reads playwright.config.ts and .spec.ts/.spec.js files from the project environment.
- Boundary markers: None explicitly mentioned to delimit code content from review instructions.
- Capability inventory: Uses the Edit tool to apply fixes, ExecutionProxy for command execution, and can spawn sub-agents.
- Sanitization: Includes a mandatory safety protocol to redact secrets and PII before sharing context with subagents.
- [COMMAND_EXECUTION]: The skill executes terminal actions via an ExecutionProxy (e.g., rtk npm test) as part of a deterministic TDD protocol to verify test failures and fixes.
Audit Metadata