cleanup-stale-code
Pass
Audited by Gen Agent Trust Hub on Jun 24, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes local scripts and binary tools including test runners, linters, formatters, and type checkers (Sections A5, B2, and B6). It also uses shell search utilities like
ripgreporgrepto scan the codebase. While these are standard for development tasks, they involve executing code and commands within the local project environment. - [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) because it ingests and processes untrusted data from the repository (code and comments) to decide on file modifications.
- Ingestion points: Codebase files are read and searched using anchors in Steps A2 and B3.
- Boundary markers: No explicit delimiters or instructions are provided to the agent to ignore embedded instructions within the processed code/comments.
- Capability inventory: The skill can perform file writes (Steps A4 and B5) and execute shell commands (linters/tests).
- Sanitization: No sanitization or validation of the ingested code content is performed before processing.
Audit Metadata