OpenAI

Install

npm install agentfootprint openai

Server-side (Node.js)

import { Agent, openai, defineTool } from 'agentfootprint';

const provider = openai('gpt-4o', {
  apiKey: process.env.OPENAI_API_KEY,
});

const agent = Agent.create({ provider })
  .system('You are a helpful assistant.')
  .tool(myTool)
  .streaming(true)
  .build();

const result = await agent.run('Hello');

Browser

No SDK dependency — calls the OpenAI REST API directly:

import { Agent, BrowserOpenAIAdapter } from 'agentfootprint';

const provider = new BrowserOpenAIAdapter({
  apiKey: userApiKey,
  model: 'gpt-4o',
});

const agent = Agent.create({ provider }).build();

Available models

Model	ID	Best for
GPT-4o	gpt-4o	General purpose
GPT-4o Mini	gpt-4o-mini	Fast, low cost
GPT-4.1	gpt-4.1	Coding, instruction following
GPT-4.1 Mini	gpt-4.1-mini	Budget coding
GPT-4.1 Nano	gpt-4.1-nano	Fastest, cheapest
o3	o3	Complex reasoning
o3 Mini	o3-mini	Fast reasoning
o4 Mini	o4-mini	Latest reasoning

Switching from the OpenAI SDK

Before (raw SDK)

import OpenAI from 'openai';

const client = new OpenAI();
const response = await client.chat.completions.create({
  model: 'gpt-4o',
  messages: [{ role: 'user', content: 'Hello' }],
});
console.log(response.choices[0].message.content);

After (agentfootprint)

import { Agent, openai } from 'agentfootprint';

const agent = Agent.create({
  provider: openai('gpt-4o'),
})
  .system('You are helpful.')
  .build();

const result = await agent.run('Hello');
console.log(result.content);
console.log(agent.getNarrative());  // full trace

Observability

Track tokens, cost, and tool usage:

import { Agent, openai } from 'agentfootprint';
import { agentObservability } from 'agentfootprint/observe';

const obs = agentObservability();

const agent = Agent.create({
  provider: openai('gpt-4o'),
})
  .system('You are helpful.')
  .recorder(obs)
  .build();

await agent.run('Hello');

console.log(obs.tokens());  // { totalCalls: 1, totalInputTokens: 18, ... }
console.log(obs.cost());    // USD estimate

Provider failover

Cross-provider resilience — the order is configurable:

import { anthropic, openai } from 'agentfootprint';
import { fallbackProvider } from 'agentfootprint/resilience';

// OpenAI primary, Claude fallback
const provider = fallbackProvider([
  openai('gpt-4o'),
  anthropic('claude-sonnet-4-20250514'),
]);

// Or Claude primary, OpenAI fallback — swap the order
const provider = fallbackProvider([
  anthropic('claude-sonnet-4-20250514'),
  openai('gpt-4o'),
]);

What you gain: Same agent code works across providers. Swap one line to switch. Add fallback for resilience.