gabrielmahia/kenya-3d

Claude + Blender MCP — civic data 3D visualization for East Africa # kenya-3d 🌍 > **Claude + Blender MCP — Civic Data 3D Visualization for East Africa** Part of the **East African Decision Infrastructure** portfolio. > ⚠️ **DEMO MODE**: Sample data files (`data/`) contain synthetic values for illustration only. > Connect live APIs via `fetchers/` for production use. See DEMO vs REAL. ## What This Does `kenya-3d` connects Claude Code to Blender via the Model Context Protocol so you can generate civic data 3D visualizations for East Africa using plain English prompts. | Visualization | Description | |---|---| | 🏙️ County Bar Chart | Per-county metrics as scaled 3D bars on Kenya's geographic footprint | | 🌋 Drought Terrain | Drought severity displaced as a terrain mesh with drought-to-red colour ramp | | 💧 Water Stress Map | County water stress as glowing emission cylinders — top-down view | ## Architecture ``` Claude Code ──► blender-mcp-server ──► Blender 4.x (bpy) │ ┌────────────┤────────────┐ │ │ │ county_bars.py drought_terrain.py water_stress_map.py │ fetchers/ wapimaji.py counties.py ``` ## Quick Start ### Prerequisites - Blender 4.x - Claude Code CLI (`npm install -g @anthropic-ai/claude-code`) - Python 3.11+ ### Setup ```bash # 1. Enable blender-mcp add-on in Blender # Edit → Preferences → Get Extensions → search "MCP" → Install & Enable # 2. Start the MCP server pip install uv uvx blender-mcp-server # 3. Connect Claude Code claude mcp add blender -- uvx blender-mcp-server claude mcp get blender # verify # 4. Install Python deps pip install -r requirements.txt ``` ### Run a Visualization With Blender open and MCP connected, tell Claude Code: ``` Load data/kenya_counties.json and run blender_scripts/county_bars.py in Blender. Colour bars by water_stress, set a 60° top-down camera, and render a preview. ``` See `prompts/` for ready-made prompt templates. ## DEMO vs REAL | Component | Mode | Notes | |---|---|---| | `data/kenya_counties.json` | ⚠️ DEMO | Synthetic metrics | …