Research-Stack/5-Applications/text-to-cad
Devin AI 0c9efac330 chore(consolidation): integrate E8Sidon stack (PRs #79 #80 #81 #89) into one PR
Squash the four overlapping feature branches into a single change set against
main, eliminating cross-PR merge conflicts and the duplicated CI-fix scripts.

What this brings in (merge order #79 -> #80 -> #81 -> #89):
- #79 refactor(infra): shared utilities (4-Infrastructure/lib/*: q16, hashing,
  jsonl, fraction_utils) + the scripts/math-first/* validators that the
  math-check CI requires.
- #80 feat(lean): Semantics.E8Sidon (1025 lines) -- Eisenstein coefficient
  identity E4^2 = E8 and the Sidon framework. E4_sq_eq_E8_coeff is fully proved
  (all Fourier-coefficient extraction machine-checked); the single residual gap
  is pinned to E4_sq_eq_E8_qExpansion (Mathlib lacks the valence formula /
  dim M8 = 1). 4 sorries + 1 axiom (e8_additive_completeness), all TODO(lean-port).
- #81 refactor(lean): Float-free FixedPoint core (integer-only sqrt/log2/expNeg).
  E8Sidon.lean kept at #80's final 1025-line version (the #81 intermediate
  438-line copy was overridden by merge order).
- #89 feat(lean): Semantics.RRC.PolyFactorIdentity -- short-sleeve polynomial
  detection at the zerocopy limb boundary; now imports Semantics.E8Sidon for
  sigma3/sigma7/convolutionLHS (single source of truth) instead of inlining them.

Conflict resolution:
- flake.nix -> canonical rs-surface removal (Garnix shutdown).
- scripts/math-first/* -> byte-identical across branches, clean.
- .cursorrules / AGENTS.md -> unified; baselines + sorry inventory refreshed.

Verification:
- lake build (default aggregator): 3573 jobs, 0 errors.
- lake build Semantics.RRC.PolyFactorIdentity (E8Sidon + FixedPoint + PolyFactor):
  3655 jobs, 0 errors. Witnesses verified (sigma7 4 = 16513, convolutionLHS 6 = 2350).
- Python tests: 68/68 pass.

Note: the "Workers Builds: researchstack" check is a preexisting external
Cloudflare build unrelated to this change (no branch touches 4-Infrastructure/cloudflare/).

Build: 3573 jobs (default), 3655 jobs (narrow), 0 errors
Co-Authored-By: Allaun Silverfox <bigdataiscoming+9i37y6j2@protonmail.com>
2026-06-16 02:01:31 +00:00
..
models initial: sovereign research stack (consolidated, weightless, and lfs-optimized) 2026-05-04 18:11:36 -05:00
skills chore(consolidation): integrate E8Sidon stack (PRs #79 #80 #81 #89) into one PR 2026-06-16 02:01:31 +00:00
viewer initial: sovereign research stack (consolidated, weightless, and lfs-optimized) 2026-05-04 18:11:36 -05:00
.gitattributes initial: sovereign research stack (consolidated, weightless, and lfs-optimized) 2026-05-04 18:11:36 -05:00
.gitignore initial: sovereign research stack (consolidated, weightless, and lfs-optimized) 2026-05-04 18:11:36 -05:00
AGENTS.md docs(agents): project-wide AGENTS.md audit — cross-refs, baseline, contracts 2026-05-26 22:34:46 -05:00
CLAUDE.md initial: sovereign research stack (consolidated, weightless, and lfs-optimized) 2026-05-04 18:11:36 -05:00
LICENSE initial: sovereign research stack (consolidated, weightless, and lfs-optimized) 2026-05-04 18:11:36 -05:00
README.md initial: sovereign research stack (consolidated, weightless, and lfs-optimized) 2026-05-04 18:11:36 -05:00
requirements-cad.txt initial: sovereign research stack (consolidated, weightless, and lfs-optimized) 2026-05-04 18:11:36 -05:00

Demo of the text-to-cad harness generating and previewing CAD geometry

⚙ Open Source Text to CAD Harness ⚙

An open source harness for generating 3D models with your favorite coding agent

GitHub stars GitHub forks License: MIT Follow @soft_servo Python build123d OCP STEP STL URDF Node.js React Vite

Features

  • Generate - Create source-controlled CAD models with coding agents like Codex and Claude Code.
  • Export - Produce STEP, STL, DXF, GLB, topology data, and URDF robot descriptions.
  • Browse - Inspect generated geometry in a local CAD Explorer viewer.
  • Reference - Copy stable @cad[...] references so agents can make precise follow-up edits.
  • Review - Render quick snapshots for fast checks during an iteration loop.
  • Reproduce - Edit source files first, then regenerate explicit targets.
  • Local - Run the harness and viewer locally with no backend to host.

🧰 Bundled Skills

This harness vendors file-targeted skills for CAD and robot-description work. Use the bundled copies here for local models/ projects, or use the dedicated repositories when installing the skills outside this harness.

🔁 Workflow

  1. Describe - Tell your agent about the part, assembly, fixture, robot, or mechanism you want.
  2. Edit - Let your coding agent update CAD source files under models/.
  3. Regenerate - Create explicit STEP, STL, DXF, GLB, or URDF targets.
  4. Inspect - Open the CAD Explorer viewer to review the generated model.
  5. Reference - Copy @cad[...] handles when you want geometry-aware edits.
  6. Commit - Save the source and generated artifacts together once the model is ready.

🚀 Quick Start

Clone the repo:

git clone https://github.com/earthtojake/text-to-cad.git
cd text-to-cad

Install Python CAD dependencies:

python3.11 -m venv .venv
./.venv/bin/python -m pip install --upgrade pip
./.venv/bin/pip install -r requirements-cad.txt

Install viewer dependencies:

cd viewer
npm install

Run the local CAD Explorer:

npm run dev

Then open http://localhost:4178.