The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses standard system utilities to perform file search and hosting operations. In commands/generate.md, it employs the find command to discover PRD files within the local project directory. In commands/preview.md, it provides instructions to initiate a local HTTP server using the python3 -m http.server command.\n- [EXTERNAL_DOWNLOADS]: The skill references the use of npx serve in commands/preview.md to provide a preview environment, which involves fetching and executing the serve package from the well-known npm registry at runtime.\n- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests untrusted text from project files and passes it to an external AI skill for code generation.\n
Ingestion points: The skill reads content from *_main_prd.md and *_module_prd.md files located in the project workspace, as specified in commands/generate.md.\n
Boundary markers: There are no explicit delimiters or instructions to ignore embedded commands within the prompt logic defined in templates/prototype_prompt_template.md.\n
Capability inventory: Capabilities include system command execution via shell (find), interaction with the frontend-design skill for file generation, and launching local network services.\n
Sanitization: No sanitization, validation, or escaping of the parsed PRD content is performed before it is interpolated into the prototype generation prompt.

prd-to-prototype