This project is an AI Agent API built using LangChain and OpenAI.
The agent can work with various tools and keeps chat history with users.

• 🧠 Works with OpenAI GPT models
• 🔧 Tool system – perform different tasks
• 💾 Stores chat history via Redis
• 🌐 RESTful API for easy integration
• 🏗️ Clean Architecture – clear and extensible code
• 📊 Token usage tracking – monitor costs

1. ⏰ Time Tool – get the current time
2. 🔍 ChromaDB Search – semantic search in PDF documents
3. 🌐 DuckDuckGo Search – internet search
4. 📚 Wikipedia Search – search on Wikipedia
5. 📡 Web Scraper – fetch data from URLs
6. 🧮 Calculator – perform mathematical operations

• Node.js 18+
• Redis server
• OpenAI API key

git clone <repository-url>
cd langchain-agent-api

npm install

Create a .env file and add the following:

# Basic settings
PORT=3000
NODE_ENV=development

# API keys
OPENAI_API_KEY="your-openai-api-key"

# Redis
REDIS_URL=redis://localhost:6379

# Agent settings
CHAT_HISTORY_TTL=604800
MODEL=gpt-4o-mini-2024-07-18
TEMPERATURE=0.7

# With Docker
docker run -d -p 6379:6379 redis:alpine

# Or locally
redis-server

# Development mode
npm run dev

# Build for production
npm run build

# Start production server
npm start

# Or with production environment
npm run start:prod

# Build Docker image
docker build -t langchain-agent-api .

# Run with Docker
docker run -p 3000:3000 --env-file .env langchain-agent-api

# Or use Docker Compose
docker-compose up -d

POST /ai/chat-with-memory/chat
Content-Type: application/json

{
  "userId": "user123",
  "input": "Hello, how can I help you?"
}

POST /ai/chat-with-memory/agent
Content-Type: application/json

{
  "userId": "user123",
  "input": "Tell me today’s weather",
  "agentType": "openai_tools",
  "enabledTools": ["duckduckgo_search", "get_current_time"]
}

GET /ai/chat-with-memory/tools

PUT /ai/chat-with-memory/tools/duckduckgo_search
Content-Type: application/json

{
  "enabled": true
}

DELETE /ai/chat-with-memory/history/user123

src/
├── agents/                     # Agents and tools
│   ├── tools/                  # Tool system
│   │   ├── base/               # Core interfaces
│   │   ├── implementations/    # Tool implementations
│   │   └── tool-registry.ts    # Tool registry
│   └── chat-with-memory/       # Chat agent
│       ├── controllers/        # HTTP controllers
│       ├── services/           # Business logic
│       ├── routes/             # API routes
│       ├── dto/                # Data transfer objects
│       └── storage/            # Data storage
├── common/                     # Common utilities
├── config/                     # Configuration
├── v1/                         # API version
├── app.ts                      # Express application
├── server.ts                   # Server entry point
└── index.ts                    # Main file

To create a new tool:

1. Create a new file in src/agents/tools/implementations/
2. Implement the IToolService interface
3. Register it in ToolRegistry

Example:

import { tool } from "@langchain/core/tools";
import { z } from "zod";
import type { IToolService } from "../base/tool.interface.js";

export class MyCustomToolService implements IToolService {
  readonly name = "my_custom_tool";
  readonly description = "My custom tool";

  createTool() {
    return tool(
      async (input: { param: string }): Promise<string> => {
        // Tool logic
        return `Result: ${input.param}`;
      },
      {
        name: this.name,
        description: this.description,
        schema: z.object({
          param: z.string()
        }),
      }
    );
  }
}

The project provides debug information via console logs:

• 🚀 Server status
• 🤖 Agent requests
• 🔧 Tool activity
• 📊 Token statistics
• ❌ Errors

Extra debugging options:

AGENT_VERBOSE=true          # Agent debug logs
NODE_ENV=development        # Development mode

1. Fork the repository
2. Create a feature branch (git checkout -b feature/new-feature)
3. Commit your changes (git commit -am 'Added new feature')
4. Push the branch (git push origin feature/new-feature)
5. Open a Pull Request

This project is distributed under the MIT License.

If you have questions or need help:

1. Create an issue on GitHub
2. Carefully read the documentation
3. Check the logs

Happy coding! 🎉