openai-agents

Files (1)

• SKILL.md

SKILL.md (first lines)

# OpenAI Agents Integration

This skill provides comprehensive guidance for integrating OpenAI Agents SDK with Temporal workflow orchestration.

## Overview

OpenAI Agents SDK allows building autonomous AI systems that can use tools, make decisions, and execute complex tasks. When combined with Temporal, you get durable, reliable execution with built-in retries and observability.

## Setup and Configuration

### 1. Installation
```bash
# Install OpenAI SDK
pip install openai

# Install with Temporal
pip install temporalio

# Create requirements.txt
cat > requirements.txt << EOF
openai>=1.0.0
temporalio>=1.0.0
python-dotenv>=1.0.0
EOF

# Install dependencies
pip install -r requirements.txt
```

### 2. Environment Variables
```python
# .env file
OPENAI_API_KEY=sk-your-api-key-here
OPENAI_MODEL=gpt-4.7
OPENAI_MAX_TOKENS=4096
OPENAI_TEMPERATURE=0.7
TEMPORAL_HOST=localhost:7233
TEMPORAL_NAMESPACE=default
TEMPORAL_TASK_QUEUE=openai-queue
```

### 3. Initialization
```python
from openai import AsyncOpenAI
from dotenv import load_dotenv
import os

load_dotenv()

client = AsyncOpenAI(
    api_key=os.getenv("OPENAI_API_KEY"),
    organization=os.getenv("OPENAI_ORG_ID", None)
)

model = os.getenv("OPENAI_MODEL", "gpt-4")
```

## Basic Usage

### Chat Completions
```python
@activity.defn
async def chat_completion(prompt: str, system_message: str | None = None) -> str:
    """Generate text using OpenAI chat completion."""
    messages = []

    if system_message:
        messages.append({"role": "system", "content": system_message})

    messages.append({"role": "user", "content": prompt})

    response = await client.chat.completions.create(
        model="gpt-4",
        messages=messages,
        max_tokens=1000,
        temperature=0.7
    )

    return response.choices[0].message.content
```

### Streaming Responses
```python
@activity.defn
async def stream_completion(prompt: str) -> str:
    """Stream OpenAI responses for real-time feedback."""
    full_response = ""

    stream = await client.chat.completions.create(
        model="gpt-4",
        messages=[{"role": "user", "content": prompt}],
        stream=True
    )

    async for chunk in stream:
        if chunk.choices[0].delta.content:
            content = chunk.choices[0].delta.content
            full_response += content
            # Send progress update
            activity.heartbeat(f"Generated: {len(full_response)} chars")

    return full_response
```

## Agents with Tools

### 1. Function Calling
```python
from typing import TypedDict

class FunctionCall(TypedDict):
    name: str
    arguments: dict

functions = [
    {
        "type": "function",
        "function": {
            "name": "get_weather",

From yasinnerten/temporal-openai-agents-sdk, Apache-2.0.

More from yasinnerten/temporal-openai-agents-sdk