What problem does it solve? Writing an accurate OpenAPI specification by hand requires manually tracing every route, middleware, validator, and data model in a codebase, which is slow and error-prone for large projects. This Skill automates the reconnaissance stage by scanning a project directory and producing structured artifacts that document the entire API surface. ## Core Features & Use Cases - Multi-language route mapping: Detects endpoints across Express, NestJS, FastAPI, Flask, Django, Spring, Gin, Laravel, Rails, Actix, and more, writing results progressively to routes.json. - Auth and middleware analysis: Identifies JWT, API key, OAuth2, Basic, and session auth schemes, per-route protection status, CORS settings, and middleware chains in auth-map.md. - Schema and validator extraction: Maps path/query parameters, request bodies, response shapes, validation libraries, and data models into schemas.md and validators.md. - Auto-scaling depth: Adjusts scan depth (fast/balanced/deep) based on file count, uses Grep/Glob first for token efficiency, and checks for an existing openapi.yaml to support merge mode. - Use Case: Point it at a legacy Express monorepo with no API documentation; it produces API_RECON.md plus intermediate artifacts that feed directly into the openapi-spec-builder stage to generate a complete openapi.yaml. ## Quick Start Run the openapi-source-recon skill on the ./src directory to map all API endpoints, authentication, and schemas into an API_RECON.md report.
