1. Error appears
2. Immediately copy entire stack trace to AI
3. Apply suggested fix without understanding
4. New error appears
5. Copy new error to AI
6. Repeat until exhausted


SyntaxError: Unexpected token '}' in JSON at position 147


DeprecationWarning: pkg_resources is deprecated as an API


Error: Cannot resolve module '@react-navigation/native'
ModuleNotFoundError: No module named 'pandas'


Error: ECONNREFUSED 127.0.0.1:5432
ConnectionError: Unable to connect to database


Error: ValidationError: "email" is required


I'm building a user registration API with Express.js and getting this validation error:

Error: ValidationError: "email" is required

Context:
- Using Joi for input validation
- Frontend is sending email in the request body
- This worked fine until I added password confirmation field
- I can see the email in console.log(req.body) before validation

What I've checked:
- Request Content-Type is application/json
- Email field is present in the request payload
- Validation schema includes email field
- Other fields validate successfully

Here's my validation schema:
[paste code snippet]

And the route handler:
[paste code snippet]

Why would Joi say email is required when it's clearly in the request?


1. Give AI specific scope: "Scan only the src/services/ folder"
2. Ask for analysis first: "What patterns do you see?" before requesting fixes
3. Review AI's understanding: "Is my understanding correct that you use JWT?" 
4. Request explanations: "Why are you suggesting this change here?"
5. Apply changes incrementally: One file at a time, test between each
6. Use version control: Always commit before letting AI scan and suggest changes


1. Read the error message completely
2. Check recent changes (git diff)
3. Verify environment/configuration
4. Review relevant logs (application logs, web server logs, database logs)
5. Try obvious fixes (restart service, clear cache, reinstall dependencies)

Success rate: ~40% of issues resolved here


6. Document what you've tried
7. Prepare context: what you're building, what changed, environment, code
8. Ask AI with full context
9. Evaluate suggestions critically
10. Understand the fix before applying

Success rate: ~50% of remaining issues resolved


11. Add extensive logging/tracing
12. Use debugger to step through code
13. Create minimal reproduction case
14. Research official documentation and community forums
15. Ask AI for specific aspects, not whole solution

Success rate: Most remaining issues resolved


## Postgres Connection Pool Exhaustion - 2025-10-03

**Error**: "remaining connection slots are reserved"
**Root Cause**: Connection pool size (10) too small for concurrent requests,
               connections not being released after queries
**Solution**: Increased pool size to 20, ensured proper connection.release()
             in finally blocks
**Learned**: Always release database connections in finally blocks,
            monitor connection pool metrics
**Time**: 45 minutes (would be 10 minutes with this knowledge)


# Check service logs
aws logs tail /aws/lambda/your-function --follow

# Verify credentials and permissions
aws sts get-caller-identity


# Docker
docker ps  # Is container running?
docker logs container-name --tail 100
docker exec -it container-name sh  # Inspect inside container

# Kubernetes
kubectl get pods
kubectl logs pod-name --tail=100
kubectl describe pod pod-name  # Check events


-- Check for long-running queries
SELECT * FROM pg_stat_activity WHERE state = 'active';

-- Analyze query performance
EXPLAIN ANALYZE SELECT ...;

-- Check connection counts
SELECT count(*) FROM pg_stat_activity;