lcs-code-review
Pass
Audited by Gen Agent Trust Hub on Aug 9, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill evaluates untrusted data such as source code and external project documentation, which constitutes an indirect prompt injection surface.\n
- Ingestion points: The workflow involves reading changed code files, diffs, and LCS artifacts like Explore, PRD, and SRS documents (SKILL.md).\n
- Boundary markers: The skill utilizes a 'Strict Chain of Truth' to maintain context integrity, though it does not implement explicit markdown delimiters for code isolation.\n
- Capability inventory: Capabilities are limited to local file system operations, execution of project-specific test suites, and producing fix instructions for downstream skills like 'lcs-task-executor' (SKILL.md).\n
- Sanitization: Not explicitly implemented; the skill relies on the agent's verification of evidence against the provided artifacts.\n- [COMMAND_EXECUTION]: The skill directs the agent to run project-specific test commands (e.g., 'npm test', 'pytest') to provide evidence for the code review verdict. This is an expected behavior for this skill's function and occurs within the project's local environment.
Audit Metadata