The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Bash tool to execute generated prototype code for testing and validation. This is an intended function of the workflow and is constrained to the project's local directory structure.
[PROMPT_INJECTION]: The skill processes user-supplied concept descriptions and local project context from CLAUDE.md to guide code generation, creating a surface for indirect prompt injection. The risk is mitigated by explicit instructions to isolate prototype code in dedicated directories and a strict prohibition on importing prototype code into production environments. * Ingestion points: [concept-description] argument and the CLAUDE.md configuration file. * Boundary markers: None explicitly defined in the instructions for input delimiters. * Capability inventory: Includes access to Bash, Write, Edit, Read, Glob, and Grep tools. * Sanitization: No explicit sanitization or validation of the input content is performed before processing.

prototype