rpg-mcp

# Azgaar Reference - Usage Scope ## What We Will NOT Reuse ### ❌ Direct Code Copying - No direct module imports from Azgaar codebase - No copy-paste of functions or classes - All code written independently from scratch ### ❌ Frontend-Specific Components - UI/rendering components (React, DOM manipulation) - Browser-based libraries (D3.js for SVG rendering) - Canvas/WebGL rendering logic - Client-side state management ### ❌ Visual Presentation Layer - SVG generation and styling - Interactive map controls - Color schemes and theming - Export/download functionality ### ❌ Browser Dependencies - IndexedDB storage - Web Workers - Browser APIs (localStorage, fetch, etc.) - Frontend build tools and configurations --- ## What We WILL Reference ### ✅ Algorithm Concepts Study algorithmic approaches for: - **Heightmap generation**: Layered noise, tectonic simulation - **Climate modeling**: Temperature gradients, moisture distribution - **Biome assignment**: Lookup tables based on climate parameters - **River generation**: Flow accumulation, drainage basins - **Region segmentation**: Territory partitioning algorithms ### ✅ Quality Gates & Validation Use as benchmarks for: - **Terrain continuity**: No abrupt elevation changes - **Biome plausibility**: Realistic climate-biome relationships - **River validity**: Downhill flow, no loops, proper branching - **Settlement placement**: Logical positioning (coasts, rivers) ### ✅ Data Structures & Formats Understand patterns for: - **Grid layouts**: Hex vs square grids, coordinate systems - **Biome tables**: Temperature/moisture matrix mappings - **Elevation ranges**: Thresholds for sea level, mountains, etc. - **River networks**: Graph structures for flow paths ### ✅ Parameter Ranges & Constants Reference realistic values for: - **Temperature gradients**: Equator-to-pole variations - **Moisture levels**: Precipitation distributions - **Elevation thresholds**: Sea level, hills, mountains - **Generation scales**: Map sizes, feature densities ### ✅ Generation Patterns Learn workflows for: - **Layered generation**: Heightmap → climate → biomes → features - **Noise functions**: Octaves, persistence, lacunarity - **Constraint satisfaction**: Ensuring geographic plausibility - **Seed-based determinism**: Reproducible generation --- ## Implementation Principles ### Independent TypeScript Implementation All code in `rpg-mcp` will be: - Written from scratch in TypeScript - Designed for Node.js server-side execution - Type-safe with strict TypeScript configuration - Schema-validated using Zod ### Deterministic & Testable Following TDD principles: - Tests written before implementation - Seedable PRNG (no `Math.random()`) - Deterministic timestamps (no `Date.now()`) - Full test coverage (positive, negative, edge cases) ### Repository Pattern Architecture Maintaining clean separation: - Schema validation at all boundaries - Database constraints matching schemas - No hidden state or side effects - Explicit input/output contracts --- ## Research Workflow 1. **Study Azgaar implementation** (reference only) 2. **Extract algorithmic concepts** (write pseudocode) 3. **Write failing tests** (TDD approach) 4. **Implement independently** (from scratch in TypeScript) 5. **Validate quality** (compare against Azgaar benchmarks) This ensures we benefit from Azgaar's research while maintaining: - Legal compliance (MIT license attribution) - Code independence (no direct copying) - Architectural alignment (server-side, schema-driven) - Quality standards (TDD, determinism, type safety)