Skip to content

QUICKSTART.md

Latest commit

 

History

History
192 lines (130 loc) · 3.79 KB

QUICKSTART.md

File metadata and controls

192 lines (130 loc) · 3.79 KB
Raw

Quick Start Guide

Get Started in 3 Minutes

Step 1: Install Dependencies

pip install -r requirements.txt

Step 2: Get OpenRouter API Key

  1. Go to https://openrouter.ai
  2. Sign up for an account
  3. Navigate to Keys
  4. Create a new API key
  5. Add some credits (minimum $1)

Step 3: Configure

# Copy the example environment file
cp .env.example .env

# Edit .env and add your key
nano .env  # or use your preferred editor

Your .env should look like:

OPENROUTER_API_KEY=sk-or-v1-...your-key-here...
OPENROUTER_MODEL=anthropic/claude-3.5-sonnet

Step 4: Validate Setup

python validate_setup.py

If you see "✓ All checks passed!" you're ready to go!

Step 5: Run Your First Agent

python examples/simple_example.py

What's Next?

Try Different Models

Edit your .env file to try different models:

# For fast responses
OPENROUTER_MODEL=anthropic/claude-3-haiku

# For best quality
OPENROUTER_MODEL=anthropic/claude-3-opus

# For OpenAI
OPENROUTER_MODEL=openai/gpt-4-turbo

# For Google
OPENROUTER_MODEL=google/gemini-pro

See all available models: https://openrouter.ai/models

Explore Examples

# Custom tools
python examples/custom_tools_example.py

# Advanced multi-tool usage
python examples/advanced_example.py

Build Your Own Agent

Create a new Python file:

from miniagent import MiniAgent, register_tool
import os
from dotenv import load_dotenv

load_dotenv()

# Define your custom tool
@register_tool
def my_tool(input_text: str) -> str:
    """Description of what your tool does"""
    return f"Processed: {input_text}"

# Create agent
agent = MiniAgent(
    model=os.getenv("OPENROUTER_MODEL"),
    api_key=os.getenv("OPENROUTER_API_KEY")
)

# Add tools
from miniagent import get_tool_description
agent.tools.append(get_tool_description(my_tool))
agent.load_builtin_tool("calculator")

# Run!
response = agent.run("Your question here")
print(response)

Common Issues

"Authentication failed"

Problem: Invalid API key

Solution:

  1. Check your .env file
  2. Make sure you copied the full key
  3. Verify key is active at https://openrouter.ai/keys

"Insufficient credits"

Problem: No credits in account

Solution: Add credits at https://openrouter.ai/credits

"Module not found"

Problem: Dependencies not installed

Solution: Run pip install -r requirements.txt

Agent not using tools

Problem: Tools not loaded or model doesn't support them

Solution:

  1. Make sure you call agent.load_builtin_tool() or add to agent.tools
  2. Try a more capable model like Claude or GPT-4

Tips

Cost Saving

Start with cheaper models for development:

  • anthropic/claude-3-haiku - Very fast, very cheap
  • openai/gpt-3.5-turbo - Good balance

Use expensive models only when needed:

  • anthropic/claude-3-opus - Most capable
  • openai/gpt-4-turbo - Very capable

Debugging

Enable verbose logging:

import logging
logging.basicConfig(level=logging.DEBUG)

Rate Limiting

Add delays between requests:

import time
response1 = agent.run("Query 1")
time.sleep(1)  # Wait 1 second
response2 = agent.run("Query 2")

Learn More

Getting Help

  1. Check the main README.md
  2. Run python validate_setup.py to diagnose issues
  3. Review examples in examples/ directory
  4. Open an issue on GitHub

Happy building! 🚀