Getting Started

Install the Axon SDK and your preferred LLM client using your package manager of choice:

npm install @memorilabs/axon

(Note: Axon supports openai, @anthropic-ai/sdk, and @google/genai as optional peer dependencies).

Quickstart Example

import 'dotenv/config';
import { OpenAI } from 'openai';
import { Axon } from '@memorilabs/axon';

// 1. Initialize the LLM Client
const client = new OpenAI({ apiKey: process.env.OPENAI_API_KEY });

// 2. Initialize Axon and Register the Client
const axon = new Axon();
axon.llm.register(client);

// 3. Register a Before Hook (e.g., logging or modifying prompts)
axon.hook.before((req, ctx) => {
  console.log(`[${ctx.traceId}] Intercepted request to model: ${req.model}`);
  // You can modify the request here before it hits the provider
  return req;
});

// 4. Register an After Hook (e.g., token usage tracking)
axon.hook.after((req, res, ctx) => {
  console.log(`[${ctx.traceId}] Received response. Tokens used: ${res.usage?.totalTokens}`);
});

async function main() {
  console.log('Sending standard LLM request...');

  // 5. Use your client exactly as you normally would!
  // Axon intercepts this call in the background.
  const response = await client.chat.completions.create({
    model: 'gpt-4o',
    messages: [{ role: 'user', content: 'What is the speed of light?' }],
  });

  console.log(`AI: ${response.choices[0].message.content}`);
}

main().catch(console.error);

Key Features

• Universal LLM Support: Works natively with OpenAI, Anthropic Claude, and Google Gemini SDKs.
• Unified Hook System: Write your logic once. Axon normalizes requests and responses across all providers so your hooks work universally.
• Zero Overhead: Plugs directly into your existing client initialization. No need to rewrite your agent or application logic.
• Streaming Support: Seamlessly handles and aggregates asynchronous streaming responses.
• Type-Safe: Written in strict TypeScript with comprehensive definitions for all provider payloads.

Supported LLMs

• OpenAI (openai)
• Anthropic Claude (@anthropic-ai/sdk)
• Google Gemini (@google/genai)

Contributing

We welcome contributions from the community! Please see our Contributing Guidelines for details on:

• Setting up your development environment
• Code style and standards
• Submitting pull requests
• Reporting issues

Support

• Discord: https://discord.gg/abD4eGym6v
• Issues: GitHub Issues

License

Apache 2.0 - see LICENSE