refactoring-patterns
Pass
Audited by Gen Agent Trust Hub on Feb 21, 2026
Risk Level: SAFE
Full Analysis
- [Indirect Prompt Injection] (SAFE): The skill is designed to process user-provided code for refactoring and includes instructions to run associated test suites. This constitutes a standard vulnerability surface for code-processing tools but is consistent with the primary purpose of the skill. 1. Ingestion points: User-provided code snippets referenced in .apm/prompts/refactor-incrementally.prompt.md. 2. Boundary markers: The prompt uses markdown code fences to delimit user content. 3. Capability inventory: The skill instructs the agent to execute test commands (e.g., 'Run FULL test suite') and basic Git operations (e.g., 'git commit') in .apm/prompts/refactor-incrementally.prompt.md. 4. Sanitization: Not explicitly present, relying on the agent's internal safety filters.
- [Prompt Injection] (SAFE): No evidence of malicious instructions, safety filter bypasses, or system prompt extraction attempts was found within the instructions or metadata.
- [Command Execution] (SAFE): The use of git commands is limited to examples for version control best practices, which is appropriate and safe within the context of a refactoring skill.
Audit Metadata