cortex-docs/cortex

Cortex - Generate interactive docs, typed SDKs from OpenAPI, AsyncAPI, GraphQL, gRPC, OpenRPC and MCP servers enriched with custom Markdown.

What it solves

Cortex simplifies the process of maintaining consistent developer tools for APIs. It eliminates the need to use separate, fragmented toolchains for generating client SDKs, interactive documentation, and AI-ready interfaces from API specifications.

How it works

Cortex uses a single project configuration file (cortex.config.yml) to ingest various API specifications—including OpenAPI, AsyncAPI, GraphQL, gRPC (Protocol Buffers), and OpenRPC—alongside Markdown files. From these sources, it automatically generates:

  • Typed SDKs: Client libraries for 11 different languages (including TypeScript, Python, Go, and Rust).
  • Interactive Documentation: Static HTML sites featuring API references and guides.
  • MCP Servers: Model Context Protocol servers that provide AI agents with typed tools, embedded specifications, and project context.

Who it’s for

It is designed for developers and API providers who need to distribute their API to both human developers (via SDKs and docs) and AI agents (via MCP servers) while keeping all three in sync from a single source of truth.

Highlights

  • Multi-Protocol Support: Handles REST, WebSockets, GraphQL, gRPC, and JSON-RPC.
  • Unified Generation: Generates SDKs, docs, and MCP servers with one command.
  • Agent-Ready: Automatically creates MCP servers that give AI agents access to API tools and documentation.
  • Broad Language Support: Supports a wide array of target languages for SDKs, from C++ and C# to Ruby and PHP.
  • Customizable: Allows output customization through Eta template overrides.

Related

  • Project
    NVIDIA-NeMo/Switchyard

    Switchyard is a pre‑1.0 NVIDIA‑NeMo project that routes each LLM request to the cheapest model that can still solve the task. It offers a Rust library, a Python package, a NeMo Relay plugin, and a demo OpenAI‑compatible proxy. Configurable routing algorithms (stage, capability, escalation, etc.) let you automatically fall back from cheap to strong models, cutting token cost by up to ~30 % while keeping accuracy near that of a single high‑end model. The code is Apache‑2.0 licensed and intended for integration into existing LLM gateways or custom harnesses.

  • Project
    eclipse-zenoh/zenoh

    A high-efficiency communication protocol that unifies publish/subscribe, geo-distributed storage, and computation for distributed systems.

  • Project
    qdrant/fastembed

    A lightweight, fast Python library for generating text, image, and multimodal embeddings using ONNX Runtime to avoid heavy PyTorch dependencies.

  • Project
    open-webui/open-terminal

    A lightweight, self-hosted terminal that provides AI agents and automation tools a sandboxed or bare-metal environment to execute code and manage files via a REST API.