cs-trick
Pass
Audited by Gen Agent Trust Hub on May 20, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a local Python utility script (
.codestable/tools/search-yaml.py) to manage and search through its document archives based on type, language, or framework filters. - [PROMPT_INJECTION]: The mandatory 'Code Investigation' phase requires the agent to ingest and analyze source code from the project repository to generate authentic examples. This represents an indirect prompt injection surface. 1. Ingestion points: Project codebase files read during Phase 2. 2. Boundary markers: Absent. 3. Capability inventory: Local file read/write access and execution of the search-yaml utility. 4. Sanitization: No explicit sanitization or filtering of codebase content is mentioned before processing.
Audit Metadata