What problem does it solve?
Choosing the right tool for 3D work is hard: mesh modeling, parametric CAD, point clouds, photogrammetry, rendering, GPU debugging, and slicing each have many providers with different costs, quality tiers, and offline requirements. This Skill provides a capability-based provider matrix so an agent can pick the right tool for each step from idea to printed part or game-ready asset.
Core Features & Use Cases
- Capability-based provider tables: Twelve capabilities (model.mesh, model.parametric, pointcloud.process, photogrammetry.reconstruct, render.offline, fabricate.slice, and more) each mapped to providers with kind, requirements, cost, quality, and offline support.
- Preflight environment checks: Probes for Python packages (trimesh, open3d, cadquery, bpy), native binaries (Blender, FreeCAD, Meshroom, PrusaSlicer, RenderDoc), and API credentials before committing to a provider.
- End-to-end recipes: Ready-made pipelines such as printable-part (CAD to gcode), game-asset (model to engine export), reality-scan (photos to clean mesh), and product-viz-still (CAD to path-traced render).
- Use Case: A user asks to turn a photo set into a 3D-printable model. The agent runs preflight, picks Meshroom for photogrammetry, CloudCompare for point cloud cleanup, trimesh for manifold checks, and PrusaSlicer for gcode — all offline and free.
Quick Start
Ask the agent to run the 3d-cad matrix preflight and then plan a printable-part pipeline from a parametric CAD model to sliced gcode.