Connect Coding Standards & Refactoring Habits

This skill defines the coding principles and refactoring patterns used throughout the Connect project. Apply these standards when writing new code, reviewing existing code, or refactoring.

Core Principles

1. Separation of Concerns

• •API Endpoints: Each endpoint should have a single, well-defined responsibility
• •Services: Business logic separated from API routes and data models
• •Models: Pure data structures with validation, no business logic
• •Frontend Components: Single responsibility - one component, one purpose
• •Example: Topic creation should be a separate endpoint (/topics/create), not mixed into chat endpoint (/chat)

2. Clean Code

• •DRY (Don't Repeat Yourself): Extract common patterns into reusable functions
• •SOLID Principles: Especially Single Responsibility and Dependency Inversion
• •No Magic Numbers/Strings: Use constants or configuration
• •Explicit over Implicit: Make intentions clear in code
• •Fail Fast: Validate inputs early, return errors immediately
• •Expressive naming: Use expressive function name rather than redundant docstring or comment

3. Expressive Naming

• •Functions: Use verb phrases that describe what they do
  • •✅ get_relevant_context() instead of get_context()
  • •✅ create_topic() instead of create()
  • •✅ merge_nodes() instead of merge()
• •Variables: Use nouns that describe what they represent
  • •✅ selectedCategoryIds instead of cats
  • •✅ memory_context instead of ctx
  • •✅ isNewSession instead of new
• •Classes: Use nouns describing the entity
  • •✅ MemoryRetriever instead of Retriever
  • •✅ GraphService instead of Service
• •Boolean variables: Use is_, has_, should_, can_ prefixes
  • •✅ isLoading, hasPermission, shouldCreateCategory

4. Simplicity

• •Avoid Over-Engineering: Solve the problem at hand, not hypothetical future problems
• •Prefer Simple Solutions: If a simple approach works, use it
• •Remove Unused Code: Delete dead code, don't comment it out
• •Minimize Dependencies: Only add dependencies when necessary
• •Clear Flow: Code should read like a story from top to bottom

5. Generalization vs Specificity

• •General Methods: Create reusable, parameterized functions when patterns repeat
  • •✅ update_node_properties(node_id, label, properties) instead of separate methods for each node type
  • •✅ get_nodes_by_label(label, filters) instead of type-specific methods
• •Specific When Needed: Don't over-generalize - keep it practical
• •Constants for Magic Values: Extract repeated strings/numbers into class constants
  • •✅ LABEL_KNOWLEDGE = "Knowledge" instead of hardcoded strings

6. Error Handling

• •Graceful Degradation: Non-critical operations (like caching) should fail silently with logging
• •Explicit Error Messages: Provide context in error messages
• •Logging Levels: Use appropriate levels (debug, info, warning, error)
• •Transaction Safety: Ensure database operations are atomic where needed

7. Code Organization

• •Logical Grouping: Group related methods together with section comments
  • •Example: # ========== Node Creation ==========
• •Consistent Patterns: Use the same patterns throughout the codebase
• •File Structure: One class per file, related utilities grouped together
• •Import Organization: Group imports (stdlib, third-party, local)

8. Type Safety

• •Type Hints: Use type hints in Python for clarity
• •Pydantic Models: Use for data validation and serialization
• •TypeScript: Use strict types, avoid any when possible
• •Optional Types: Use | None or ? to indicate nullable values

9. Documentation

• •Docstrings: Use for public functions/classes explaining what and why
• •Comments: Explain "why" not "what" - code should be self-documenting
• •README: Keep updated with architecture decisions
• •Inline Comments: Only when code intent isn't obvious

10. Testing Considerations

• •Testable Code: Write code that's easy to test (dependency injection, pure functions)
• •Edge Cases: Consider edge cases during implementation
• •Integration Tests: Test full flows, not just units

Refactoring Patterns

When Refactoring:

1. •Identify Duplication: Look for repeated patterns
2. •Extract Common Logic: Create general helper methods
3. •Simplify Conditionals: Use early returns, guard clauses
4. •Remove Dead Code: Delete unused functions/variables
5. •Rename for Clarity: Improve names to match current understanding
6. •Consolidate Similar Operations: Merge related functions

Refactoring Checklist:

• • Are concerns properly separated?
• • Are function/variable names expressive?
• • Is the code simple and readable?
• • Are there any hardcoded values that should be constants?
• • Is there duplication that can be extracted?
• • Are error cases handled gracefully?
• • Is the code organized logically?
• • Are types properly annotated?

Project-Specific Patterns

Backend (Python/FastAPI)

• •Use async/await consistently for I/O operations
• •Use dependency injection (Depends()) for services
• •Group endpoints logically in main.py
• •Use Pydantic models for request/response validation
• •Use logger from logging module, not print()

Frontend (React/TypeScript)

• •Use functional components with hooks
• •Extract reusable logic into custom hooks
• •Use TypeScript interfaces for props and data structures
• •Use TanStack Query for data fetching and caching
• •Use Tailwind CSS utility classes, avoid inline styles
• •Keep components small and focused

Database Operations

• •Use transactions for multi-step operations
• •Handle connection errors gracefully
• •Use parameterized queries (prevent SQL injection)
• •Batch operations when possible (Redis pipeline, Neo4j batch)

Anti-Patterns to Avoid

• •❌ Mixing business logic in API endpoints
• •❌ Creating endpoints that do multiple unrelated things
• •❌ Using generic names like data, item, result without context
• •❌ Hardcoding values that might change
• •❌ Creating overly specific methods when a general one would work
• •❌ Ignoring errors silently without logging
• •❌ Premature optimization
• •❌ Over-abstracting simple operations

Example: Good vs Bad

Bad:

async def chat(request: ChatRequest):
    session_id = request.session_id or str(uuid.uuid4())
    session = await session_manager.get_session(session_id)
    if not session:
        # Creates topic, creates category, creates sub-categories, etc.
        # 50+ lines of mixed concerns
    # Then handles chat logic

Good:

@app.post("/topics/create")
async def create_topic(request: CreateTopicRequest):
    # Only handles topic creation
    ...

@app.post("/chat")
async def chat(request: ChatRequest):
    # Only handles chat messaging
    ...

When to Use This Skill

• •When writing new code - ensure it follows these standards
• •When reviewing code - check against these principles
• •When refactoring - use these patterns to improve code quality
• •When debugging - consider if code organization could be improved
• •When adding features - ensure proper separation of concerns

Instructions

1. •Before Writing Code: Review relevant sections of this skill
2. •While Coding: Apply naming conventions, separation of concerns, and simplicity principles
3. •After Coding: Review against the refactoring checklist
4. •When Refactoring: Identify patterns, extract common logic, improve names
5. •Ask Questions: Use the ask questions tool if you need to clarify requirements with the user

Remember

• •Code is read more than written - prioritize readability
• •Simple solutions are often the best solutions
• •Names are the most important form of documentation
• •Separation of concerns makes code maintainable
• •Refactor incrementally - small improvements add up