nuke-code
Pass
Audited by Gen Agent Trust Hub on Aug 15, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill facilitates the execution of standard repository maintenance tools such as linters, type-checkers, and test suites (e.g., npm, pytest, cargo). These commands are resolved from local project manifests (package.json, pyproject.toml) to ensure code quality before finalizing tasks.
- [PROMPT_INJECTION]: The instructions establish a strict 'constitution' and verification protocol that prevents self-verification by spawning a separate validator in a fresh context. This design is specifically intended to mitigate bias and ensure objective review of code changes.
- [DATA_EXFILTRATION]: The skill includes a dedicated security lens and checklist that instructs the agent to actively hunt for and prevent the inclusion of secrets, API tokens, and credentials in the codebase. No unauthorized external network operations were identified.
- [REMOTE_CODE_EXECUTION]: No patterns of downloading and executing untrusted scripts or payloads were found. The skill relies entirely on established local development environments and tools.
Audit Metadata