The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from multiple documentation files to generate a sequential build roadmap, creating an attack surface where instructions embedded in project files could influence the agent's output.
Ingestion points: The skill reads ./docs/PRD.md, ./docs/TECH_STACK.md, progress.txt, ./docs/BACKEND_STRUCTURE.md, ./docs/FRONTEND_GUIDELINES.md, and ./docs/APP_FLOW.md.
Boundary markers: There are no explicit delimiters or instructions provided to the agent to disregard embedded prompts within the ingested files.
Capability inventory: The skill is designed to write the final roadmap to ./docs/IMPLEMENTATION_PLAN.md.
Sanitization: The skill does not perform validation or sanitization of the content read from the documentation files before using it in the output.
[EXTERNAL_DOWNLOADS]: The skill generates instructions for project initialization and dependency installation using official registries and well-known tools.
Evidence: The generated plan includes commands like npx create-next-app, npm create vite, and standard pip install commands for frameworks such as FastAPI, Flask, and Django.
Vendor resources: The skill references vendor-owned tools including vibe-coding-css-setup, vibe-coding-local-runner, and vibe-coding-ship for project configuration and deployment tasks.

vibe-coding-doc-implplan