CadQuery

1. Overview

Python library for scripted parametric CAD whose models are ordinary code.

Category: Engineering & CAD.

2. Version and scope assessed

Version assessed: Source commit 33361dc1b5b70bc83ce36f1b34a01ca85dfca80a; STEP fixture, assembly exporter, and CQGI execution source inspected; no CAD execution.

4. Principle profile

5. Fundamental requirements

6. Evidence and verification

• E1: Resolved: the pinned STEP fixture is readable ISO-10303-21 text with blue_cylinder and red_cube product names, two MANIFOLD_SOLID_BREP entities, and nine COLOUR_RGB records. No geometry was reconstructed, and the sample is not evidence of an export executed in this review. (source), confidence high
• E2: Source inspection shows the assembly STEP writer explicitly enables colour, layer, and name modes. Retaining names and colour is therefore supported by this code path; constraints, editable intent, and receiving-tool fidelity still require comparison. (source), confidence medium
• E3: Source code is independently readable, but a clean environment rebuild with pinned CadQuery and OCCT dependencies was not performed. (source), confidence medium
• E4: Model generation and export run in a local Python environment without a required hosted service. (source), confidence medium
• E5: Resolved: CQGI builds an environment with real Python builtins, compiles the input model, and calls exec; exceptions are captured in the build result. The model is executable code. This source inspection does not establish a sandbox or actual authorization and overwrite safeguards. (source), confidence medium
• sovereignty: A positive Sovereignty finding requires established E1, E2 and E4 in the declared scope. The existing access/exit/custody record for CadQuery does not establish all three minimums. The prior positive finding is withdrawn; recorded storage and portability strengths remain facts, not a verified failure. (source), confidence medium
• longevity: E3 is not established for CadQuery's existing essential-record scope. The previous Longevity finding is replaced by Not assessed, not a verified failure or a positive placeholder. (source), confidence medium
• locality: The documented workflow is local and scriptable. (source), confidence medium
• transparency: The positive control observations above do not yet establish whole-scope E5. The methodology requires that minimum before a positive Transparency and Control finding; the previous positive finding is therefore withdrawn to Not assessed, not changed to Does not meet. (source), confidence medium
• E3: The STEP fixture exposes product names, solids and colors, and assembly export enables names/layers/colors. Geometry equivalence and the editable source's constraints, parameters and OCCT/runtime dependencies have not been reconstructed. Readable Python plus STEP entity counts do not establish the declared engineering-intent recovery. (source), confidence medium
• E5: CQGI executes model text with real Python builtins and captures exceptions; the inspected path does not establish a sandbox. No effective execution authorization, dependency trust, output-overwrite visibility or correction after model side effects is established for the declared workflow. (source), confidence medium
• E2: STEP name/color/layer output does not establish migration of the essential parametric constraints and editable intent into an independent CAD workflow. The former E2 Pass is therefore unestablished, not a verified failure of every possible exit route. (source), confidence medium

7. Migration and recovery

Native storage format: Python scripts that export STEP and STL.

Synchronization or runtime model: Local Python environment; plain-text scripts suit version control.

Example export or inspection command (not verified against any particular version or data set): python -c "import cadquery as cq; cq.exporters.export(cq.Workplane().box(1,1,1), 'box.step')"

8. Architectural strengths

Established strengths and their scope are recorded in the evidence and assessment findings on this page.

9. Architectural limitations

• STEP metadata is inspectable, but geometry and parametric constraints were not reconstructed. CQGI executes Python with real builtins.

10. Assessment findings and remaining work

Assessment update, 23 October 2026:

• Resolved: STEP product names and colour records are independently readable; the inspected assembly exporter enables name, layer, and colour modes.
• Resolved: CQGI executes model Python with real builtins and captures exceptions in its build result.
• Open: Geometry equivalence, editable constraints in another CAD tool, complete offline rebuild, and execution/overwrite safeguards. E5 needs model/dependency execution authority, overwrite controls and correction boundaries; exception capture is not isolation.

Assessment: Direct sample inspection establishes access to STEP metadata. Source inspection establishes export options and the executable-model boundary. Full engineering recovery and consequential oversight remain Not Verified; the overall determination remains Insufficient evidence.

Human-readable criteria: Can an independent party preserve the source model, units, dependencies, and editable structure, reproduce the result, and exchange it without losing important design meaning?

Rating reconciliation: 24 October 2026

E3 and E5 remain Not Verified for the existing declared scope. Longevity and Transparency and Control are Not assessed pending their required minimums. Sovereignty is also Not assessed because the required access/exit/custody minimums are not all established. The prior E2 Pass is withdrawn; its essential-record migration comparison remains incomplete. This is a targeted evidence decision, not a finding that the product fails. Existing E1, E4 and Locality findings were not re-certified.

Remaining: Geometry equivalence, editable constraints in another CAD tool, complete offline rebuild, and execution/overwrite safeguards. E5 needs model/dependency execution authority, overwrite controls and correction boundaries; exception capture is not isolation.

Next without installing locally: Inspect source, pinned dependency records and STEP/native output with expected geometry/constraint manifests supplied by upstream or an assessor. Review execution/overwrite safeguards without running untrusted model code.

The individual decision and provenance are retained in evidence/rating-reconciliation/batch-5/. Overall: Insufficient evidence.

Individually reconciled in rating batch 5 on 24 October 2026. E3/E5 remain Not Verified for explicit missing coverage; dependent unsupported findings corrected. Previously resolved questions are retained, not counted again. E1/E4/Locality were not re-certified by this targeted review. The assessment procedure is described in the methodology.