simplify-code
Pass
Audited by Gen Agent Trust Hub on Jul 5, 2026
Risk Level: SAFE
Full Analysis
- [DATA_EXPOSURE]: The skill instructs the agent to inspect local source code and git history to identify refactoring targets. This is standard developer behavior and does not involve exfiltrating data to external sources.
- [COMMAND_EXECUTION]: The workflow involves running local verification tools like formatters, linters, and test runners. These are standard development utilities used to ensure code quality after modifications.
- [INDIRECT_PROMPT_INJECTION]: While the skill processes untrusted code, the instructions provide clear boundaries to prevent unintended behavior changes. 1. Ingestion points: local files and git output (SKILL.md Step 1). 2. Boundary markers: absent. 3. Capability inventory: file modification (SKILL.md Step 3) and local command execution (SKILL.md Step 5). 4. Sanitization: absent. However, the instructions specifically mandate that security checks, validation, and existing contracts remain untouched.
Audit Metadata