How yukionishi1129/techpicks works
Project type and purpose
• This is a full-stack, microservice-based monorepo for “Check Picks.”
• It comprises multiple frontend applications (web client v1, web client v2, admin panel, React Native mobile stub), a GraphQL BFF (Apollo Gateway), several Go-based gRPC microservices, batch jobs, and a Supabase (PostgreSQL) database.
File structure
• Top-level directories:
• web/client-v2, web/admin, web/client – Next.js frontends
• mobile/rn-client – React Native mobile app
• bff/apollo-gateway – NestJS GraphQL gateway
• micro-service/… – content-service, bookmark-service, favorite-service, my-feed-service, user-service (Go + gRPC)
• batch-service – Go batch jobs (crawlers, seed/migrations, trend crawlers)
• supabase – DB migrations & config
• Each microservice follows a clean architecture: internal/application, internal/adapter, internal/domain, infrastructure, interfaces (handlers).
• Configuration/build files: Dockerfiles, cloudbuild.yaml, Makefiles, .env.sample.
Key README insights
• Frontends use Next.js, TypeScript, Apollo Client, supabase-js, Tailwind.
• BFF uses NestJS + Apollo Server to stitch GraphQL schemas, delegates resolvers to underlying gRPC services.
• Microservices are Go binaries with gRPC servers, using SQLBoiler for ORM, deployed to Cloud Run.
• batch-service runs as Cloud Run Jobs for periodic crawls (RSS, 3rd-party APIs) and Firestore seeding.
Main components
a. Frontends (Client-v2, Admin, optional Mobile)
b. BFF / GraphQL Gateway
c. Microservices (gRPC servers):
– Content Service
– Bookmark Service
– My-Feed Service
– Favorite Service
– User Service
d. Batch Processing Jobs (batch-service)
e. Database layer: Supabase (PostgreSQL)
f. External APIs & data sources: RSS feeds, OGP, 3rd-party article APIs
g. CI/CD pipelines & infrastructure (Cloud Run, Cloud Build)
Relationships and interactions
• User browser/mobile → HTTP/HTTPS → Next.js frontend
• Next.js → GraphQL over HTTP → Apollo Gateway
• Apollo Gateway → gRPC → Microservices
• Microservices → SQLBoiler → Supabase (PostgreSQL)
• Batch-service jobs → external APIs (RSS, site crawlers) → Firestore or Supabase
• Microservices may call each other via gRPC or via Gateway
• External authentication via Supabase auth or JWT
Architectural patterns & design principles
• Microservices with clear domain boundaries, gRPC for inter-service comms
• GraphQL gateway for unified API endpoint and schema stitching
• Clean architecture within Go services (ports & adapters)
• Next.js ISR/SSR frontends decoupled from backend
• Batch jobs decoupled from real-time services
Final checklist
• Components: Frontends, Gateway, Microservices, Batch Jobs, Database, External APIs
• Interactions: HTTP, GraphQL, gRPC, ORM calls, external API pulls
• Labels: component name + protocol/tech (e.g., “Bookmark Service – gRPC”)
• Deployment boundaries: Cloud Run, Supabase, local/dev