Generate Your AI Agent in Minutes with OpenAPI

Easily create a powerful LangChain based AI agent using your OpenAPI specification no manual coding required. Just enter your OpenAPI URL or upload a JSON/YAML file, and instantly generate a fully functional agent equipped with dynamic tools and intelligent task orchestration.

Dynamic API Toolset

This platform automatically builds intelligent tools and agents from your OpenAPI spec, transforming each API endpoint into callable components the agent can use to interact with external services.

Agent Architecture

• Agents: Each agent is assigned a specific task domain and is equipped with a custom toolset derived from the relevant API endpoints. This allows the agent to act autonomously and efficiently within its scope.
• Supervisor: A central supervisor agent oversees the workflow. It evaluates the current task state and delegates responsibilities to the appropriate agents, ensuring smooth execution and avoiding redundancy.
• State Graph: The system uses a state graph to coordinate agent actions, track progress, and route communication between agents and the supervisor, resulting in a cohesive and structured task flow.

LLM Integration

All agents are powered by a large language model (LLM), enabling them to:

• Interpret natural language instructions
• Make intelligent decisions
• Execute multi-step API calls
• Communicate outcomes effectively

Start Now

• Step 1: Paste your OpenAPI URL or upload the spec file
• Step 2: Click "Generate Agent"
• Step 3: Run and explore your intelligent API-powered agent

API Agent Guide: Build an AI Agent with OpenAPI

Quick Start

In this guide, you'll learn how to create an AI agent that interacts with the JSONPlaceholder API using its OpenAPI specification.

Resources

• API Documentation: https://jsonplaceholder.typicode.com/guide/
• OpenAPI Spec (JSON): Download Spec

Step-by-Step Instructions

1. Generate Your Agent Project

1. Visit: https://agentsapis.com/
2. Upload the JSON OpenAPI spec you downloaded.
3. Download the generated .zip file.
4. Unzip and open the folder in your preferred code editor.

2. Set Up the Project

Inside the unzipped folder:

Add Your OpenAI API Key

Create or update the .env file with your API key:

Install Dependencies

Run one of the following commands:

3. Configure the API Client

Open src/index.ts and set your API URL to https://jsonplaceholder.typicode.com.

4. Run the Script

Execute the script using ts-node:

You should see output from the agent calling the list posts endpoint using OpenAPI and LLM inference.

Explore the Core Logic

Once your agent is running, dive into src/core.ts to see how it all works under the hood:

• LLM-driven graph creation

• Tool-to-endpoint mapping

• Execution logic via Axios

You're Done!

You’ve successfully built an AI-powered agent that uses OpenAPI to intelligently query an external REST API. You can now build on top of it by customizing tools, modifying prompts, or adding your own logic.

Try different prompts like:

• "Get all users"

• "Create a new comment for post 1"

• "Delete a todo item"

Happy building!