Skip to content

langchain-openapi-tools

Convert OpenAPI specifications into native, production-grade LangChain tools for AI agents.

langchain_openapi automatically parses OpenAPI v3.0 and v3.1 specifications and generates dynamic, strongly typed LangChain StructuredTools. No code generation or manual tool writing is required.

Key Features

  • ⚡ Zero-Code Tool Generation: Instant runtime conversion from OpenAPI JSON/YAML to LangChain tools.
  • 🔒 Comprehensive Authentication: Built-in support for Bearer Tokens, API Keys, Basic Auth, Cookies, and custom request providers.
  • 🛡️ Production Middleware Pipeline: Onion-style request/response interceptors for retries, rate-limiting, caching, logging, and result pagination.
  • 🎯 Type-Safe Validation: Dynamic Pydantic schema generation ensures LLM inputs are validated before HTTP execution.
  • 🌐 Async & Sync Support: Fully asynchronous execution engine powered by httpx.

Quick Example

from langchain_openapi_tools import OpenAPIToolkit

# Load spec from remote URL or local file
toolkit = OpenAPIToolkit.from_url("https://api.crossref.org/swagger-docs")

# Get dynamically generated LangChain tools
tools = toolkit.get_tools()

print(f"Generated {len(tools)} tools:")
for tool in tools:
    print(f"- {tool.name}: {tool.description}")