Skip to main content

​
Git workflow

DualMind Lab uses a simple branch-based workflow across all three repositories.

​
Branch strategy

BranchPurpose
mainProduction-ready code, deployed automatically
feature/*New features (branch from main)
fix/*Bug fixes
devIntegration branch (optional)

​
Commit conventions

Use descriptive commit messages: 
feat: add streaming support for dual chat
fix: resolve FK violation on thread creation
refactor: extract model selection into service
docs: update API reference for voting endpoint

​
Testing

​
Backend tests

Located at tests/DualMind.API.Tests/: 
cd DualMind_Back
dotnet test
Test files:
  • SmokeTests.cs — basic endpoint connectivity
  • UnitTest1.cs — unit tests

​
Frontend linting

cd DualMind_UI
npm run lint        # Check for issues
npm run lint:fix    # Auto-fix

​
Manual API testing

Use the Swagger UI at /swagger or curl: 
# Health check (no auth)
curl http://localhost:5079/health

# Single chat (requires JWT)
curl -X POST http://localhost:5079/api/arena/chat \
  -H "Authorization: Bearer YOUR_JWT" \
  -H "Content-Type: application/json" \
  -d '{"prompt": "Hello, world!"}'

# Dual chat
curl -X POST http://localhost:5079/api/arena/dualchat \
  -H "Authorization: Bearer YOUR_JWT" \
  -H "Content-Type: application/json" \
  -d '{"prompt": "Explain AI", "selectionMode": "random"}'

​
Deployment process

​
Backend (Azure)

The backend deploys to Azure App Service via GitHub Actions: 
# Triggers on push to main
# Builds .NET 8 app
# Deploys to Azure App Service

​
Frontend (Cloudflare Workers)

cd DualMind_UI
npx wrangler deploy

​
Admin panel (Cloudflare Workers)

cd DualMind_Admin-UI
npm run deploy       # staging
npm run deploy:prod  # production

​
Adding a new API endpoint

  1. Create or update a controller in Controllers/
  2. Add route attribute: [Route("api/your-route")]
  3. Add HTTP method attribute: [HttpGet], [HttpPost], etc.
  4. Add [Authorize] if auth required, [AllowAnonymous] if public
  5. Inject required services via constructor
  6. Add request/response models in Core/Models/ if needed
  7. Register any new services in Program.cs
  8. Test via Swagger or curl
  9. Update OpenAPI spec in docs/openapi.yaml

​
Adding a new AI provider

  1. Create a new service class implementing IChatProvider in AI/Providers/
  2. Implement ChatAsync() and StreamAsync() methods
  3. Register the typed HttpClient in Program.cs
  4. Add the provider case in ChatProviderFactory.GetProvider()
  5. Add provider record to providers table in Supabase
  6. Add API keys to provider_api_keys table
  7. Add models using this provider to ai_models table