neutree-ai/openapi-to-skills

What it solves

AI agents often struggle with large OpenAPI specifications because they can exceed LLM context limits or waste tokens by loading unnecessary information. This tool converts monolithic OpenAPI files into a structured, modular format that allows agents to read only the specific documentation they need on-demand.

How it works

It takes an OpenAPI specification (from a local file or URL) and breaks it down into a hierarchical set of small, focused markdown files. This structure organizes content by resources, operations, and schemas. Because these files are standard markdown, any AI agent capable of reading files can navigate the documentation by starting with an overview and drilling down into specific sections as required.

Who it’s for

Developers building AI agents who want to provide their models with access to complex API documentation without sacrificing performance or hitting context window limits.

Highlights

• Modular output: Splits massive API specs into thousands of small, navigable files.
• Context efficiency: Enables agents to load only relevant API operations or schemas rather than the entire specification.
• Flexible filtering: Allows users to include or exclude specific tags, paths, or deprecated operations.
• Customizable: Supports custom output templates via Eta templates and provides a programmatic API for integration into build pipelines.