refactor-code
Pass
Audited by Gen Agent Trust Hub on Jun 22, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill focuses on standard software engineering practices for refactoring Elixir code. It emphasizes reversible steps, test-driven verification, and maintaining stable public interfaces.
- [COMMAND_EXECUTION]: The instructions mandate the use of
mix test, which is the standard, local build and test tool for Elixir projects. This is a legitimate and expected operation for the skill's stated purpose. - [PROMPT_INJECTION]: The skill uses authoritative language (e.g., "HARD-GATE", "NEVER", "MUST") to enforce engineering safety and consistency. These instructions are directed at maintaining code quality and do not attempt to bypass the AI's core safety guidelines or instructions.
- [INDIRECT_PROMPT_INJECTION]: The skill establishes an attack surface by ingesting and modifying user-provided Elixir source code. However, it incorporates significant mitigation by requiring the agent to write and run local characterization tests before and after any changes, ensuring that the behavior remains consistent and that the agent relies on actual execution results ("Observed output") rather than fabricated or injected instructions.
Audit Metadata