AgentSkillsCN

Backend Architect

后端架构师

SKILL.md

Backend Architect Skill - Architecture Intelligence

Searchable database of backend architectures, databases, security guidelines, language best practices, API patterns, and platform engineering recommendations.

Prerequisites

  • Python 3.8+
  • No external dependencies (uses only standard library)

How to Use This Skill

Step 1: Analyze User Requirements

Extract key information from user request:

  • Product type: SaaS, E-commerce, Fintech, Social, Chat, IoT, AI Agent, etc.
  • Scale requirements: Startup (<20 devs), Enterprise (>50 devs), Global scale
  • Critical features: Real-time, High-write, ACID transactions, AI/ML integration
  • Preferred stack: Go, Python, Node, Java, .NET, Rust

Step 2: Search Relevant Domains

Use search.py multiple times to gather comprehensive information:

bash
python3 .gemini/skills/backend-architect/scripts/search.py "<keyword>" --domain <domain> [-n <max_results>]

Recommended search order:

  1. Product - Get architecture recommendations for product type
  2. Architecture - Get detailed architecture pattern (components, trade-offs)
  3. Database - Get database recommendations for data model
  4. Security - Get security guidelines (OWASP, API Security)
  5. Language - Get language/framework recommendations
  6. API - Get API pattern recommendations (REST, GraphQL, gRPC)
  7. Stack - Get stack-specific guidelines

Step 3: Stack Guidelines

bash
python3 .gemini/skills/backend-architect/scripts/search.py "<keyword>" --stack <stack>

Available stacks: go, python, node, java, dotnet, rust

Step 4: Verify Latest Versions (STRICT)

Before starting work, always verify the latest stable versions of technologies via web search. AI training data has a knowledge cutoff (e.g., Go 1.24 is now out, but AI might only know 1.23).

  • Search Query: latest stable version <technology> 2026
  • What to check: Language runtimes (Go, Python, Node), Base Docker images, Primary frameworks.
  • Where to apply: Dockerfile, go.mod, package.json, requirements.txt.

Example: If AI suggests Go 1.23 but 1.24 is out, use 1.24-alpine in Dockerfile.


Search Reference

Available Domains

DomainUse ForExample Keywords
architectureSystem architecture patternsmicroservices, modular monolith, serverless
databaseDatabase selectionpostgresql, mongodb, redis, vector
securitySecurity best practicesOWASP, injection, auth, encryption
productProduct-specific recommendationsecommerce, saas, fintech, chat
languageLanguage/framework selectionrust, go, python, node
apiAPI design patternsrest, graphql, grpc, websocket
namingNaming conventionspython, go, sql, graphql
errorError handling patternshttp, grpc, postgres
platformDevOps/Platform toolskubernetes, terraform, argocd

Available Stacks

StackFocus
goGin/Echo, Cloud-native
pythonFastAPI/Django, AI/ML
nodeNestJS/Fastify, Full-stack
javaSpring Boot, Enterprise
dotnetASP.NET Core, Cloud-native
rustAxum/Actix, High-performance

Example Workflow

User request: "Build a real-time chat application"

bash
python3 .gemini/skills/backend-architect/scripts/search.py "chat messaging" --domain product
python3 .gemini/skills/backend-architect/scripts/search.py "real-time event driven" --domain architecture
python3 .gemini/skills/backend-architect/scripts/search.py "high write cassandra" --domain database
python3 .gemini/skills/backend-architect/scripts/search.py "websocket" --domain api

Common Rules for Professional Backend

Database Selection

  • OLTP: PostgreSQL, MySQL
  • High-write: Cassandra, ClickHouse
  • Caching: Redis
  • Vector (AI): pgvector, Qdrant

Architecture Decision

  • Team < 20: Modular Monolith
  • Team > 50: Microservices
  • Real-time: Event-Driven
  • AI/LLM: Agentic Architecture