Agentlify JavaScript/TypeScript Client

The Serverless Infrastructure for AI Agents.

Build, run, and scale production AI agents with a single SDK. Agentlify handles the infrastructure—routing, observability, rate limits, and tool execution—so you can focus on building features.

Why Agentlify?

Think of us as "Netlify for LLMs":

• DevX First: Install one SDK, access any model (OpenAI, Anthropic, Google, etc.).
• Serverless Agents: Define tools with simple callbacks in your existing code (Next.js API routes, Node.js services). No need for a separate "agent server" or complex Python microservices.
• Smart Infrastructure: Automatic model routing (cost/speed/quality), retries, and fallbacks.
• Full Observability: Real-time logs, cost tracking, and tool execution traces out of the box.

Features

• Unified Agent API: Build agents that run anywhere.
• Local Tool Callbacks: Execute tools in your code, secure and simple.
• Smart Model Routing: Automatically optimize for cost, speed, or quality.
• Multi-Provider Support: Switch models without changing code.
• Function Calling: First-class support for OpenAI-compatible tools.
• Streaming: Built-in support for real-time responses.
• TypeScript: Full type safety included.

Installation

npm install agentlify-js
# or
yarn add agentlify-js

🚀 Quick Start: Building an Agent

Create an agent that can interact with your local code or APIs using tools with callbacks.

const Agentlify = require('agentlify-js');

const client = new Agentlify({
  apiKey: process.env.AGENTLIFY_API_KEY,
  routerId: process.env.AGENTLIFY_ROUTER_ID,
});

// Run an agent with a local tool
const response = await client.agents.run({
  agentId: 'my-agent', // Create in Agentlify dashboard
  messages: [{ role: 'user', content: 'What is the stock price of AAPL?' }],
  tools: [
    {
      type: 'function',
      function: {
        name: 'get_stock_price',
        description: 'Get current stock price',
        parameters: {
          type: 'object',
          properties: { symbol: { type: 'string' } },
          required: ['symbol'],
        },
      },
      // This callback executes in YOUR code
      callback: async (args) => {
        // Call your DB, external API, etc.
        const price = await fetchStockAPI(args.symbol);
        return { price, currency: 'USD' };
      },
    },
  ],
});

console.log(response.choices[0].message.content);

📚 Core Concepts

1. Agents & Tools

Agents are the core of Agentlify. You can define them in the dashboard and run them via the SDK. Tools can be Webhooks (server-side) or Callbacks (local).

// Example: Database Query Tool
const response = await client.agents.run({
  agentId: 'data-assistant',
  messages: [{ role: 'user', content: 'Find users in New York' }],
  tools: [
    {
      type: 'function',
      function: {
        name: 'query_db',
        description: 'Query database',
        parameters: {
          type: 'object',
          properties: { city: { type: 'string' } },
        },
      },
      callback: async (args) => {
        // Securely access your DB here
        return await db.users.find({ city: args.city });
      },
    },
  ],
});

2. Smart Routing

Stop hardcoding models. Let Agentlify route to the best model for the task.

const completion = await client.chat.create({
  messages: [{ role: 'user', content: 'Analyze this complex contract...' }],
  // Agentlify selects GPT-4, Claude 3 Opus, or others based on your router config
});

3. Streaming

Stream responses to your frontend easily.

const stream = await client.chat.create({
  messages: [{ role: 'user', content: 'Write a long story...' }],
  stream: true,
});

for await (const chunk of stream) {
  process.stdout.write(chunk.choices[0]?.delta?.content || '');
}

⚙️ Configuration

const mp = new Agentlify({
  apiKey: 'your-api-key',
  routerId: 'YOUR_ROUTER_ID',
  timeout: 30000,
});

📖 Documentation

For full documentation, visit docs.agentlify.co.

• Agent Tools Guide
• SDK Reference
• Router Configuration

🤝 Support

• 📧 Email: help@agentlify.co
• � Issues: GitHub Issues

📝 License

MIT License - see LICENSE file for details.