We built a lightweight Typescript library that turns GraphQL APIs into tool definitions for LLMs like GPT, Claude, and others. It also integrates directly with agentic frameworks like LangChain.

To build agents and chatbots, we found ourselves building custom tooling and APIs to pull in information, outsource analytics, and trigger external actions. But wiring that up, doing schema validation, data mapping, retry loops, etc entails tedious boilerplate code. We generalized that code into a library, released it under OSS Apache 2.0, and hope that it helps others build GenAI apps that need to connect data sources, trigger actions, or outsource symbolic reasoning.

Would love your feedback and thoughts. Happy to share why we settled on GraphQL for the API (validation, semantic annotations, simple mapping, query flexibility) and also curious what you are using and why.