refactor-safely
Pass
Audited by Gen Agent Trust Hub on May 26, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill defines a structured workflow for code refactoring with a strong emphasis on safety protocols, including preservation boundaries and characterization tests.
- [PROMPT_INJECTION]: No instructions were found that attempt to override system behavior, bypass safety guardrails, or leak system prompts. The instructions use standard imperative language to guide the agent through a refactoring process.
- [DATA_EXFILTRATION]: There are no network-related commands (curl, wget) or attempts to access sensitive system files like credentials or SSH keys.
- [REMOTE_CODE_EXECUTION]: The skill does not perform any external downloads or execute code from remote sources. It references internal framework skills (e.g.,
framework:clean-code), which are standard for modular agent architectures. - [COMMAND_EXECUTION]: The skill involves code modification as part of its primary purpose (refactoring), but it includes significant guardrails such as 'Zero Refactor Rules' and mandatory user approval gates before any edits occur.
Audit Metadata