@ruah-dev/conv

Converter

How do I make this API agent-sized?

Curate and convert massive API specifications into compact, agent-friendly tool surfaces. Compress 400-endpoint specs into 5–10 task-level MCP tools with minimal context overhead.

Source

See it

@ruah-dev/conv

Quick start

Up and running in seconds

@ruah-dev/conv
$ ruah conv curate api.yaml
# 29 endpoints  10 tools
# ~226 tok of definitions (was ~540)

$ ruah conv generate api.yaml --curate --json

Features

015 input formats (OpenAPI 3.x, Swagger 2.0, Postman, GraphQL, HAR)
02Smart endpoint curation (collapses bloated APIs into task-level tools)
03MCP Tool Definitions & TypeScript/Python server scaffolds
04OpenAI & Anthropic Function Calling schema generation
05Context footprint & token cost calculator per toolset
06Single runtime dependency

5

Input formats

≤10

Curated tools

1

Dependency

Command Reference

Commands

ruah conv curate <spec>Interactively collapse endpoints into ≤10 task tools
ruah conv generate <spec>Parse specification and generate target tool surface
ruah conv inspect <spec>Parse and display intermediate representation (IR) summary
ruah conv validate <spec>Validate API spec against best practices for agents
ruah conv targetsList all supported output surface targets

Input

Supported formats

OpenAPI 3.x
Swagger 2.0
Postman v2.1
GraphQL SDL
HAR

Output

Target surfaces

MCP Tool Definitions
MCP Server (TypeScript)
MCP Server (Python)
OpenAI Function Calling
Anthropic Function Calling
A2A Service

Dive deeper

Full API reference, guides, and examples.