dead-code-sweep
Pass
Audited by Gen Agent Trust Hub on Feb 22, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION] (LOW): The skill executes build and linting commands (e.g., 'npm run build') defined in the codebase configuration during Phase 4. While intended for verification, this could trigger malicious scripts if the target codebase is untrusted.
- [PROMPT_INJECTION] (LOW): The skill is susceptible to indirect prompt injection because it processes untrusted code. 1. Ingestion points: The entire codebase is read during detection phases. 2. Boundary markers: No specific delimiters or safety instructions are provided to the sub-agents analyzing file contents. 3. Capability inventory: The skill can delete files, modify code, and execute shell commands. 4. Sanitization: The skill implements a 'human-in-the-loop' safeguard by requiring user approval via AskUserQuestion before any cleanup occurs.
Audit Metadata