sweep
Pass
Audited by Gen Agent Trust Hub on Oct 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the local repository that could contain malicious instructions designed to trick the LLM-based analysis components.
- Ingestion points: Scans source files, comments, and documentation across the project as defined in
reference/cleanup-protocol.md. - Capability inventory: Generates deletion plans that are handed off to the
Builderagent for execution as specified inSKILL.md. - Boundary markers: The skill does not explicitly define delimiters or specific "ignore" instructions for content passed to the LLM during the "DCE-LLM" detection pattern.
- Sanitization: No explicit sanitization or escaping of file content is described before it is analyzed by the AI.
- [COMMAND_EXECUTION]: The skill executes standard development and static analysis tools to identify cleanup candidates.
- Evidence: Utilizes tools such as
knip,vulture,staticcheck, andclippyto perform reachability analysis across different languages. - [EXTERNAL_DOWNLOADS]: The skill uses official package managers to run analysis tools, which may involve fetching dependencies from public registries.
- Evidence: Invokes
npx knipfor TypeScript analysis, which may download the package from the npm registry if it is not present in the local environment.
Audit Metadata