cs-learn
Pass
Audited by Gen Agent Trust Hub on May 20, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill runs a local utility
python .codestable/tools/search-yaml.pyto identify duplicate entries in the project's learning database. This is a functional capability used for repository maintenance.\n- [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it ingests and processes untrusted user input and existing documents to generate new project learning files.\n - Ingestion points: User dialogue context and existing project-specific markdown files in the
.codestable/directory.\n - Boundary markers: None; the skill does not use specific delimiters or instructions to ignore embedded commands within ingested data.\n
- Capability inventory: The skill can perform file system writes within the local
.codestable/directory and execute project-local Python scripts.\n - Sanitization: Not present; the drafted content is based on raw input without explicit escaping or validation steps.
Audit Metadata