What problem does it solve? Before technical design begins, teams often lack organized information about the external APIs, SDKs, and services a requirement depends on, causing design interruptions and missed integration details. This Skill scans requirements for third-party dependencies and produces one structured document per dependency, stored in the requirement directory. ## Core Features & Use Cases - Dependency Identification: Scans requirement descriptions to detect third-party APIs, SDKs, external services, and data sources, then presents a checklist for user confirmation. - Parallel Information Collection: When two or more dependencies exist, it dispatches sub-agents via task-dispatch to collect documentation in parallel, reusing or building web indexes to avoid redundant site exploration. - Structured Output: Generates a README index plus one Markdown file per dependency covering interfaces, authentication, rate limits, alternatives, and risks, with optional demo verification scripts that read secrets from environment variables. - Use Case: A requirement involves OpenAI API and Stripe payments. The Skill identifies both, collects their docs in parallel, and writes dependencies/README.md, openai-api.md, and stripe.md into the requirement directory for the design phase to reference. ## Quick Start Ask the agent to organize the third-party dependency documentation for the current requirement before starting technical design.
