Add CodingBuddy Configuration Schema

[JeremyDev87]

Add CodingBuddy Configuration Schema

📋 Summary

Introduces a standardized configuration schema for CodingBuddy that allows users to provide project-specific context (tech stack, architecture, conventions, test strategy) to AI assistants. All fields are optional to lower the barrier to entry.

Closes #19

🎯 Problem

Context Gap

AI coding assistants don't know your project context. When developers start using AI assistants, they must repeatedly explain:

• What language they want responses in
• What tech stack they're using
• What architectural patterns they follow
• What coding conventions their team prefers

This creates a "repetition tax" where developers re-explain project context in every conversation.

Current Limitations

• No standardized way to communicate project context
• AI makes assumptions that may not fit the project
• Inconsistent responses across different AI tools
• Team members get different AI behavior

✨ Solution

Configuration Schema

A comprehensive, optional configuration schema that provides project context:

1. Basic Settings

   • language: Response language (ko, en, ja, etc.)
   • projectName: Project name
   • description: Project description
   • repository: Repository URL

2. Tech Stack (techStack)

   • Languages, frontend, backend, database, infrastructure, tools
   • Optional deep details (version, config, notes)

3. Architecture (architecture)

   • Pattern (layered, hexagonal, clean, feature-sliced, etc.)
   • Structure (directory paths)
   • Component style (flat, grouped, feature-based)
   • Optional layer definitions

4. Conventions (conventions)

   • Style guide (airbnb, google, standard, prettier)
   • Naming conventions (files, components, functions, etc.)
   • Import order, line length, semicolons, quotes
   • Optional custom rules

5. Test Strategy (testStrategy)

   • Approach (tdd, bdd, test-after, mixed)
   • Frameworks, coverage target
   • Unit test pattern, E2E directory
   • Mocking strategy

6. Additional Context

   • keyFiles: Important files AI should know about
   • avoid: Topics or areas to avoid
   • custom: Freeform custom context

Key Design Principles

• All fields optional: Start with what you need
• No rule override: Provides context only; .ai-rules remain unchanged
• Type-safe: TypeScript types inferred from Zod schemas
• Validated: Runtime validation with clear error messages
• Extensible: Easy to add new configuration items

📊 Files Changed

• New Files (5):
  • docs/config-schema.md (266 lines) - Comprehensive documentation
  • mcp-server/src/config/config.types.ts (43 lines) - TypeScript types
  • mcp-server/src/config/config.schema.ts (202 lines) - Zod validation schema
  • mcp-server/src/config/config.schema.spec.ts (273 lines) - Test suite
  • mcp-server/src/config/index.ts (34 lines) - Public API exports

Total: 5 files changed, +818 insertions

🔑 Key Features

Type Safety

• TypeScript types inferred from Zod schemas (single source of truth)
• Full IDE autocomplete support
• Type guards for runtime validation

Validation

• Zod schema validation with clear error messages
• Validates enum values, number ranges, URL formats
• Provides error paths for nested validation failures

Test Coverage

• Comprehensive test suite covering:
  • Empty config (all fields optional)
  • Minimal config
  • Full config with all fields
  • Deep optional fields
  • Invalid values and edge cases
  • Type inference verification

Documentation

• Complete schema reference with examples
• Quick start guide
• TypeScript usage examples
• Validation error examples

✅ Benefits

1. Consistency: All AI assistants use the same configuration format
2. Explicitness: AI responds based on clear information rather than guessing
3. Flexibility: Only specify the settings you need
4. Extensibility: Structure makes it easy to add new configuration items
5. Type Safety: Full TypeScript support with autocomplete

📝 Usage Example

// codingbuddy.config.js
module.exports = {
  language: 'ko',
  projectName: 'my-awesome-app',
  techStack: {
    frontend: ['React', 'TypeScript'],
    backend: ['NestJS'],
  },
  architecture: {
    pattern: 'feature-sliced',
    structure: ['src/app', 'src/features', 'src/shared'],
  },
  conventions: {
    style: 'airbnb',
    naming: {
      files: 'kebab-case',
      components: 'PascalCase',
    },
  },
  testStrategy: {
    approach: 'tdd',
    frameworks: ['vitest'],
    coverage: 80,
  },
};

🧪 Testing

• Empty config validation (all fields optional)
• Minimal config validation
• Full config validation
• Deep optional fields validation
• Invalid enum values rejection
• Invalid number ranges rejection
• Type inference verification
• Error path reporting