An automated check suite
for STEP assemblies.

Geometric checks, BOM-vs-CAD audit, and an honesty toolchain for STEP-based CAD. Like pytest for mechanical design — in spirit; CADCLAW reports findings with severity and a confidence budget rather than just pass/fail.

Install CADCLAW View on GitHub

Open source · MIT · Python 3.10+ · No commercial CAD software required for CADCLAW's own checks

53interferences caught

70 → 13 MBSTEP size reduction

150+validated design changes

13check modules

Numbers from the M3-CRETE deployment. CADCLAW catches problems your config tells it to look for; it doesn't certify designs.

CAD assemblies break silently.

Parts clip into each other. BOMs drift from geometry. A motor mount ends up 600 mm from the motor. Engineers catch these by eye — if they catch them at all. There is no pytest for CAD.

CADCLAW validates STEP assemblies through a chain of automated gates. A harness run passes only when every configured check passes.

Inventory

Missing or extra parts. Labels by bounding-box signature, counts against expected. Per-region (axis-aligned) constraints supported.

Interference

Solid-solid overlaps via BRep boolean intersection — not just bbox.

Adjacency

Parts that should be near each other but aren't. Catches the motor 600 mm from its mount.

Dimensional

Wrong thickness, swapped box() args, impossible dimensions.

Orientation

Expected face planes for labels that opt in through cadclaw.yaml.

Floating

Non-exempt parts isolated from configured structural labels beyond a max gap.

Color/material

STEP AP242 color metadata checked against expected label colors.

Kinematics

Beam deflection, motor torque budgets, belt tension, racking.

Tolerance

Worst-case, RSS, Monte Carlo stacking with C_pk and variance decomposition.

BOM audit

BOM JSON ↔ CAD assembly: qty, mfg_type, required/forbidden text terms, CAD-side count drift. Private fields never echoed.

Parity

STEP-vs-STEP comparison; flags hidden/suppressed-part export drift (file shrinks but unique signatures grow).

Disassembly

Sequenced part removal, radial exploded views, animation frame export.

Render

STEP → PNG → animated GIF via offscreen VTK. Closes the loop to shareable visuals.

Honesty toolchain: cadclaw doctor verifies your environment before MCP errors get opaque · cadclaw publish-audit stops private build BOMs from being committed · cadclaw claim-audit flags overclaims and untagged numeric assertions in your README and BOM notes. Configure once in cadclaw.yaml, run from the cadclaw CLI.

MCP Server included. Core validation, analysis, and audit modules are exposed as MCP tools so an MCP-compatible assistant or client can call them directly. The MCP server only exposes CADCLAW's own checks; it does not give the assistant access to your CAD application or to anything outside what cadclaw itself can do.

What CADCLAW does NOT prove

CADCLAW checks the geometry of a STEP file, the JSON of a BOM, and the text of your README against rules you write. It does not prove:

That the native CAD model has no hidden or suppressed parts. CADCLAW reads the STEP export, which can silently drop invisible parts.
That the physical build matches the CAD. CAD passing CADCLAW says nothing about whether the parts on your bench match the file.
That a vendor part is in stock, available, or the price you assumed.
That a printed part is strong enough for production use. CADCLAW's kinematics gates do bare-beam math; they don't simulate printed-PLA fatigue, layer adhesion, or thermal creep.
That a structural claim is physically certified, unless you've attached measurement data with an evidence tag.
That an AI-generated CAD change is correct without passing the gates. CADCLAW is the check; not passing it doesn't make a change correct, only "passed the gates we have."

Each report includes a confidence budget per gate: checked, not_checked, assumptions. Read it.

How it works

Every solid in a STEP file has a bounding box. The sorted dimensions (dx, dy, dz) rounded to 0.1 mm form a signature — a fingerprint that identifies part types without needing part names or metadata.

(40.0, 80.0, 1000.0) → "beam"     # 4080 C-beam extrusion
(56.4, 56.4, 76.6)   → "motor"    # NEMA23 stepper
(4.0, 80.0, 96.0)    → "mount"    # motor mount plate

This works because mechanical parts have characteristic dimensions. A NEMA23 is always 56.4 mm square. A 4080 extrusion is always 40×80 mm. CADCLAW exploits this invariant to label, count, and validate without parsing STEP metadata.

Sample output

The render gate closes the loop from STEP validation to shareable visuals. One function call produces a radial explode, 360° orbit, and animated GIF — the same tool you run in CI.

99-part STEP assembly exploding radially and rotating 360 degrees

99-part M3-CRETE assembly · render_radial_explode_gif("assembly.step", "out.gif") · zero manual animation

Quick start

pip install cadclaw
cadclaw doctor                # 1. verify the environment

CLI workflow

Configure once in cadclaw.yaml — labels, expected inventory, regions, BOM rules, claim-audit terms, publish-audit globs — then drive everything from the cadclaw console script:

cadclaw harness --rules cadclaw.yaml             # run configured YAML-backed checks
cadclaw bom-audit --rules cadclaw.yaml           # or run a single gate
cadclaw publish-audit --rules cadclaw.yaml       # before `git push`
cadclaw claim-audit --rules cadclaw.yaml --report-format md -o report.md

Exit codes: 0 pass · 1 fail · 2 warn-only · 3 internal error.

BOM-vs-CAD audit (the v0.6 headline)

# cadclaw.yaml fragment
bom_audit:
  bom_path: bom/data.json
  rules:
    - id: 5
      expected_qty: 12
      expected_label: connector_bar
      forbidden_terms: ["maximum rigidity", "primary stiffness"]
    - id: 65
      expected_qty: 3
      expected_unit: "bars (1.0m each)"
      expected_mfg_type: buy
      required_terms: ["1m", "friction-fit"]
      forbidden_terms: ["JB Weld", "West System", "custom 2m cut"]

Catches BOM qty / mfg_type / unit mismatches, required-term-missing or forbidden-term-present in the description, CAD-side count drift, and BOM items with no CAD geometry (suppressed for mfg_type: consumable / electronic / fastener). Private BOM fields (vendors, sku, unit_cost) are dropped at the serializer and never appear in any report.

Programmatic API

from cadclaw.harness import Harness
from cadclaw.adjacency import AdjacencyRule

h = Harness("my_assembly.step")

h.add_inventory(
    labels={(40.0, 80.0, 1000.0): 'beam', (56.4, 56.4, 76.6): 'motor'},
    expected={'beam': 4, 'motor': 2, 'belt': 3}
)
h.add_interference(skip_labels={'belt', 'wheel'})
h.add_adjacency(rules=[
    AdjacencyRule('motor', 'bracket', max_distance=50)
])

report = h.run()
print(report)

CI/CD in any system

# .github/workflows/cad-check.yml
- name: Validate assembly
  run: |
    pip install cadclaw
    cadclaw doctor
    cadclaw harness --rules cadclaw.yaml

Exit code 0 = passed. Exit code 1 = failed. Exit code 2 = warn-only.

Who it's for

Open-source hardware projects

Catch assembly errors before builders hit them.

CadQuery / FreeCAD users

The testing layer the open CAD ecosystem is missing.

Small manufacturing teams

Automated QA between design and procurement.

AI-assisted CAD workflows

Validate that AI-generated changes don't break the assembly.

Origin

CADCLAW was developed alongside the M3-CRETE open-source concrete 3D printer — built out of a practical need to properly position and validate components during assembly of a large, part-dense machine. Across that project the harness:

Caught 53 solid-solid interferences in a single run
Reduced STEP file size from 70 MB to 13 MB by identifying geometry bloat
Validated 150+ assembly changes across 15 design sessions without visual inspection
Prevented 3 regressions that would have shipped broken geometry to builders

Snapshot from CADCLAW's first production deployment on M3-CRETE, Q1 2026. Developed by Sunnyday Technologies.

Citation

If you use CADCLAW in published research or derivative work, please cite:

Sonnentag, N. (2026). CADCLAW: Automated validation framework for
STEP-based CAD assemblies. Sunnyday Technologies.
https://github.com/sunnyday-technologies/CADCLAW
DOI: 10.5281/zenodo.19647391

A CITATION.cff file is included for automated citation tooling.