I Spoke to My API — Before It Was Even Built
summary
Uploads an OpenAPI spec, auto-validates and enriches the YAML, and exposes the API as a chat-driven simulator using a custom backend and OpenAI agents. The system spins up a per-endpoint tool via an OpenAPI agent, runs POST/GET/PATCH calls against an intelligent smart mock that generates schema-valid data and maintains state, and can diff and inject new operations into the spec on demand. Architecture includes an MCP server for memory and tool routing; limitations include hallucination risk and incomplete integration with full multi-service backends; roadmap adds externalizable design rules for standards and linting (executable OpenAPI).
AI-written summary · about
chapters (5)
- 0:00API design challenges
- 1:52Traditional API development problems
- 3:52API design and testing tool
- 5:42Live demo of tool
- 14:53Q&A and future specs
related talks
26mMocking and Deploying an AI Chatbot in Minutes
29mBeyond OpenAPI and AsyncAPI: AI agents wrote the spec, code, and tests - now w…
40mAPI Lifecycle: Smarter Design and Governance With Open Tooling
28mOpey, the AI for API Discovery and Future of Fintech
28mIs Your API Agentic-AI Ready?
19mAutomate your API to perfection
35mAPI-First! Contract-First! Make API development great again!
19mSecure and accelerate APIs' lifecycle with the OpenAPI specification