Allow LLMs to explore large [OpenAPI](https://www.openapis.org/) schemas without bloating the context.