There’s no need for an adapter with the OpenAI SDK since all agentic tools are compatible with OpenAI by default. You can use AIFunctionSet.specs for function calling or AIFunctionSet.toolSpecs for parallel tool calling.

​
Install

npm install @agentic/stdlib openai

​
Usage

import { WeatherClient } from '@agentic/stdlib'
import OpenAI from 'openai'

const weather = new WeatherClient()
const openai = new OpenAI()

const messages: OpenAI.ChatCompletionMessageParam[] = [
  {
    role: 'system',
    content: 'You are a helpful assistant. Be as concise as possible.'
  },
  { role: 'user', content: 'What is the weather in San Francisco?' }
]

const res = await openai.chat.completions.create({
  messages,
  model: 'gpt-4o-mini',
  temperature: 0,
  tools: weather.functions.toolSpecs,
  tool_choice: 'required'
})
const message = res.choices[0]?.message!
console.log(JSON.stringify(message, null, 2))
assert(message.tool_calls?.[0]?.function?.name === 'get_current_weather')

const fn = weather.functions.get('get_current_weather')!

const toolArgs = message.tool_calls[0].function.arguments
const toolResult = await fn(toolArgs)
console.log(JSON.stringify(toolResult, null, 2))

​
Running this example

You’ll need a free API key from weatherapi.com to run this example. Store it in a local .env file as WEATHER_API_KEY.

You’ll need an OpenAI API key to run this example. Store it in a local .env file as OPENAI_API_KEY.

git clone git@github.com:transitive-bullshit/agentic.git
cd agentic
pnpm install
echo 'WEATHER_API_KEY=your-key' >> .env
echo 'OPENAI_API_KEY=your-key' >> .env
npx tsx examples/openai/bin/weather.ts

​
OpenAI Responses API

Agentic also supports the OpenAI Responses API by referencing AIFunctionSet.responsesToolSpecs as in this example:

import 'dotenv/config'

import type { ResponseInput } from 'openai/resources/responses/responses.mjs'
import { WeatherClient } from '@agentic/stdlib'
import OpenAI from 'openai'

async function main() {
  const weather = new WeatherClient()
  const openai = new OpenAI()

  const messages: ResponseInput = [
    {
      role: 'system',
      content: 'You are a helpful assistant. Be as concise as possible.'
    },
    { role: 'user', content: 'What is the weather in San Francisco?' }
  ]

  {
    // Call to OpenAI to invoke the weather tool
    const res = await openai.responses.create({
      model: 'gpt-4o-mini',
      temperature: 0,
      tools: weather.functions.responsesToolSpecs,
      tool_choice: 'required',
      input: messages
    })

    const message = res.output[0]
    console.log(JSON.stringify(message, null, 2))
    assert(message?.type === 'function_call')
    assert(message.name === 'get_current_weather')

    const fn = weather.functions.get('get_current_weather')!
    const toolResult = await fn(message.arguments)

    messages.push(message)
    messages.push({
      type: 'function_call_output',
      call_id: message.call_id,
      output: JSON.stringify(toolResult)
    })
  }
}

await main()

​
Running this example

You’ll need a free API key from weatherapi.com to run this example. Store it in a local .env file as WEATHER_API_KEY.

You’ll need an OpenAI API key to run this example. Store it in a local .env file as OPENAI_API_KEY.

git clone git@github.com:transitive-bullshit/agentic.git
cd agentic
pnpm install
echo 'WEATHER_API_KEY=your-key' >> .env
echo 'OPENAI_API_KEY=your-key' >> .env
npx tsx examples/openai/bin/weather-responses.ts