Troubleshooting
Quick fixes for the most common issues you’ll encounter during setup, development, and deployment.
Quick Reference
| Issue | Fix |
|---|---|
| Git hooks not running | Run npm install to reinstall Husky hooks |
| AI ignoring project patterns | Be explicit: “Follow AGENTS.md” in your prompts |
| Port conflicts | See Port Conflicts below |
| Tests failing | Restart services and re-run tests |
| Prisma errors after changes | Regenerate client and run migrations |
| Build fails in CI/CD | Test locally first with npm run build |
| Peer dependency conflicts | Install with --legacy-peer-deps flag |
| Cache/build issues | Clear cache and rebuild |
AI Framework Issues
Git Hooks Not Working
Problem: Commits succeed but no documentation is generated.
Fix: Run npm install to reinstall Husky hooks. Hooks are managed by Husky and auto-install via npm.
For more on how auto-documentation works, see Keeping Docs Synced.
AI Ignoring Project Patterns
Problem: AI generates generic code and doesn’t follow your project conventions.
Fix: Be explicit in your prompts — plan mode alone doesn’t guarantee the AI will follow patterns:
- “Follow AGENTS.md”
- “Check that scoped rules loaded for testing and security”
- “Use patterns from existing code in the dashboard”
For effective prompt patterns, see Prompt Library & Skills.
Setup Issues
General rule: If something goes wrong during setup, clean up before retrying:
make setup-cleanup # Remove all resources
make setup # Start fresh
Authentication Issues
Symptoms: “Not authenticated” or “credentials expired” errors.
Re-authenticate with the relevant services and retry setup. Both Google Cloud and GitHub CLI credentials may need refreshing.
Missing Prerequisites
Symptoms: “command not found” errors for gcloud, gh, or docker.
Ensure all prerequisites are installed:
| Tool | Version | Purpose |
|---|---|---|
| Node.js | 18+ | Runtime |
| Docker | Latest | Local database |
| GitHub CLI | Latest | Repository management |
| gcloud CLI | Latest | GCP deployment (optional) |
Port Conflicts
Symptoms: “Port 5432 already in use” or “Port 3000 already in use.”
Check what’s using the port and stop conflicting services. The setup wizard can handle this automatically in most cases.
Development Issues
Common Runtime Errors
| Error | Fix |
|---|---|
| ”Cannot find module” | Run npm install |
| ”Prisma Client not found” | Run npx prisma generate |
| ”Database not found” | Start dev server (includes database) |
| “Clerk keys missing” | Check environment file has all Clerk variables |
| TypeScript errors | Run npx tsc --noEmit to see all errors |
| Stale build | Clear cache and rebuild |
Docker Issues
If Docker containers are stale or not running, clean them up and restart the dev server. Make sure Docker Desktop is running before starting development.
Prisma Errors
After schema changes, regenerate the Prisma client and run pending migrations. If the database is in an inconsistent state, a reset may be needed (this deletes data).
Tests Timing Out
E2E tests can take over a minute. Tell your AI tool to run tests in background mode to avoid timeout issues.
Deployment Issues
Build Fails in GitHub Actions
Test the build locally first before pushing. Fix any errors, then push again.
Cloud Run Deployment Fails
Common causes:
- Missing secrets — verify all required secrets are configured
- IAM permissions — ensure the service account has the correct roles
- Container crashes — check Cloud Run logs for runtime errors
Production Feature Issues
| Issue | Fix |
|---|---|
| Emails not sending | Register your app with Inngest Cloud |
| Team invitations fail | Redeploy to auto-configure the app URL |
| Stripe redirects to wrong URL | Redeploy to auto-configure the app URL |
| Background jobs not running | Register app URL in Inngest Cloud |
For Inngest setup, see Background Jobs. For Stripe configuration, see Billing.
Getting Help
Collect Diagnostics
Run version checks for Node.js, npm, Docker, and your CLI tools. Share the output with your AI assistant for targeted help:
"I'm having [issue]. Here are my diagnostics: [paste version output]"
Additional Resources
- Setup Guide — Installation and configuration
- Prompt Library — Effective AI prompts
- Build Your First Feature — Framework walkthrough
Ready to build with VibeReady?
Get the full AI-native SaaS foundation with production infrastructure, AI development framework, and all integrations.
Get VibeReady — From $149