Quick Start (5 Minutes)
Get AgentMap running in 5 minutes with this streamlined guide.
⚡ Progress Tracker
Total Time: ~5 minutes
- Step 1: Install (30 seconds)
- Step 2: Set API Key (30 seconds)
- Step 3: Create workflow (2 minutes)
- Step 4: Run workflow (30 seconds)
- Step 5: Verify success (30 seconds)
Step 1: Install AgentMap
pip install agentmap
Expected time: 30 seconds
Step 2: Set Your API Key
Pick one provider and set your API key:
# Option A: OpenAI (recommended for beginners)
export OPENAI_API_KEY="your-api-key-here"
# Option B: Anthropic (Claude)
export ANTHROPIC_API_KEY="your-api-key-here"
# Option C: Google (Gemini)
export GOOGLE_API_KEY="your-api-key-here"
Getting API Keys
- OpenAI: Get key from platform.openai.com/api-keys
- Anthropic: Get key from console.anthropic.com
- Google: Get key from ai.google.dev
Expected time: 30 seconds
Step 3: Create Your First Workflow
Create a file called hello_world.csv
:
graph_name,node_name,agent_type,next_node,on_failure,prompt,input_fields,output_field
HelloWorld,Start,input,PrintResult,HandleError,"Hello! What is your name?",,name
HelloWorld,PrintResult,echo,,,"Hello {name}. Welcome to AgentMap!",name,result
HelloWorld,HandleError,echo,,,Error occurred
Expected time: 2 minutes
Step 4: Run Your Workflow
✨ New Simplified Syntax (Recommended):
agentmap run --csv hello_world.csv
Traditional Syntax (Still Supported):
agentmap run --csv hello_world.csv --graph HelloWorld
Custom Graph Names:
# Override graph name using :: syntax
agentmap run --csv hello_world.csv::MyCustomBot
Expected time: 30 seconds
Step 5: Success! 🎉
You should see:
Hello! What is your name?
> [Enter your name, e.g., "Alex"]
Hello Alex. Welcome to AgentMap!
Expected time: 30 seconds
⚠️ Quick Troubleshooting
Issue: "No LLM providers available"
Problem: API key not set correctly.
Solution:
# Check if your key is set
echo $OPENAI_API_KEY
# If empty, set it properly:
export OPENAI_API_KEY="your-actual-key-here"
Issue: "Module not found"
Problem: AgentMap not installed properly.
Solution:
# Reinstall with verbose output
pip install --upgrade agentmap
agentmap diagnose
Issue: "CSV parsing error"
Problem: CSV file format issue.
Solution: Copy the exact CSV content from Step 3 above. Ensure no extra spaces or special characters.
✅ You're Ready!
Congratulations! You just:
- ✅ Installed AgentMap
- ✅ Set up AI provider access
- ✅ Created a multi-agent workflow in CSV
- ✅ Ran your first AI workflow
🚀 Next Steps (Choose Your Path)
🏗️ Build Your First Real Workflow
First Workflow Guide →
15-minute tutorial: Document analysis workflow
First Workflow Guide →
15-minute tutorial: Document analysis workflow
💡 Pro Tips
- Start Simple: Master single-agent workflows before building complex chains
- Test Incrementally: Add one agent at a time when building larger workflows
- Use Templates: Check out examples for common patterns
- Get Help: Join discussions at GitHub
Having issues? See the full Installation Guide for detailed troubleshooting.