Manual Setup
Add monitoring to your AI agents manually. This guide shows you exactly where to put the decorators and what they do.
Prefer automatic setup? Use our CLI Setup for automatic code generation.
What you’re doing
You’re adding monitoring decorators to your main agent functions - the functions that handle complete user requests. Everything inside these functions gets captured automatically.
Step 1: Install and configure
Step 2: Add monitoring to your agent
Where to put it: On your main agent function that handles user requests.
What NOT to do
Don’t add decorators to helper functions:
# ❌ Don't do this - these get captured automatically
@tracing(agent="helper")
async def classify_intent(message):
return await llm.classify(message)
# âś… Just your main agent function needs the decorator
@tracing(agent="customer-service")
async def process_customer_request(message):
intent = await classify_intent(message) # This gets captured automatically
return await generate_response(intent) # This gets captured automatically
Verify it’s working
- Run your agent with a test request
- Check dashboard.handit.ai - you should see traces appearing
- Look for your agent name in the dashboard
Done! Your autonomous engineer can now see your AI and will start monitoring for issues.
Next steps
Set up evaluation: Add Quality Assessment so your engineer can detect quality issues.
Enable fixes: Connect GitHub Integration so your engineer can create pull requests with improvements.