Pioneers boundary representation (BUILD at Cambridge) and constructive solid geometry so computers distinguish solid interior from air.
Classical Roadblock: Proprietary CAD files could not exchange exact surfaces across vendors.Engineering Toolbox · Code-First Parametric CAD & STEP AP242 · X.02 (TOOL-ENG-1) · 22 min read
Code-First Parametric Solids & Measured STEP AP242, 3MF, and GLB Round-Trip Fidelity
Empowering solo mechanical and civil engineers to generate exact B-rep solids from version-controlled Python scripts and measure exactly what survives—and what is lost—across STEP AP242, 3MF, and GLB.
Independent engineers and small hardware shops shouldn't need $5,000-per-seat proprietary CAD licenses just to version-control a bracket, clevis, or structural node in Git. Code-first CAD in Python lets you parameterize every dimension as an equation, diff changes in pull requests, and export STEP AP242, 3MF, and GLB with a mathematical audit of chordal tessellation error.
1-Click Neutral Geometry Downloads
Download the committed, SHA-256-verified 3D mesh, STEP solid, 3MF print package, or DXF cut sheet directly:
Works with the Tools You Already Use
Open .glb in Blender 4.2+ (File → Import → glTF 2.0), .step in FreeCAD 1.0 or build123d / CadQuery, .3mf in OrcaSlicer / PrusaSlicer, or .dxf in QCAD Community Edition / LibreCAD / Inkscape.
Generator + Byte-Level Verifier
Run python skills/toolchain/x02/make_x02.py to generate custom dimensions and python skills/toolchain/x02/verify_x02_toolchain_x02_parametric_step_3mf_roundtrip.py to verify watertightness in <1s.
About this illustration
This AI-generated illustration is imagining what the concept would look like: an artistic impression, not a measured structure or specification. Exact 3D geometry, binary buffers and governing equations are kept strictly in deterministic Toolbox X.02 and Python generatorskills/toolchain/x02/make_x02.py below.
ES-MX Resumen en español · X.02: Sólidos Paramétricos en Código y Fidelidad Medida entre STEP AP242, 3MF y GLB
Actos I y II — Historia, ley geométrica e intuición cotidiana: Sólidos Paramétricos en Código y Fidelidad Medida entre STEP AP242, 3MF y GLB — Empoderando a ingenieros mecánicos y civiles independientes para generar sólidos B-rep exactos desde scripts en Python bajo control de versiones y medir qué se conserva y qué se pierde entre STEP AP242, 3MF y GLB. Este monográfico interactivo verifica la ley geométrica h_sagitta = R (1 − cos(π / N)) · V_step = V_flange + V_boss = 43,200 mm³ · ΔV / V < 10⁻⁶. Por qué importa en el estudio y taller: Los ingenieros independientes y talleres pequeños no deberían necesitar licencias propietarias de $5,000 USD anuales solo para controlar versiones de una pieza o nodo estructural en Git. El CAD basado en código Python permite parametrizar cada dimensión, revisar diffs en pull requests y exportar STEP AP242, 3MF y GLB auditando el error cordal de teselación.
En Palabras Sencillas (Explicación Accesible): Diseñar una pieza mecánica mediante código paramétrico es como escribir una receta clara en lugar de hacer un dibujo estático. En lugar de trazar líneas fijas, definimos dimensiones mediante fórmulas matemáticas. Por ejemplo, indicamos que la base mide sesenta milímetros de largo y que el soporte superior se ubica justo en el centro. Si más adelante cambiamos una medida, toda la pieza se actualiza de forma automática. El verdadero reto surge al compartir el archivo con talleres y colaboradores. Algunos sistemas requieren superficies analíticas perfectamente curvas, mientras que otros programas convierten todo en pequeños triángulos planos. Al transformar un orificio circular en caras planas, el diámetro efectivo se reduce y los pasadores metálicos pueden atascarse. Nuestra herramienta interactiva en el navegador les permite explorar un pedestal escalonado, ajustar el número de divisiones circulares y observar cómo cambia la redondez en tiempo real. En pocos segundos, pueden descargar modelos verificados y rutinas en Python para auditar sus propias piezas sin pagar licencias comerciales.
Mecanismo Técnico, Fuerzas en Competencia y Umbral Crítico: El intercambio de modelos geométricos entre formatos neutros ilustra la tensión matemática entre representaciones de frontera exactas y aproximaciones discretizadas por teselación. El estándar ISO 10303-21 STEP AP242 preserva definiciones topológicas analíticas mediante entidades MANIFOLD_SOLID_BREP compuestas por bucles cerrados y caras avanzadas, asegurando que el volumen teórico permanezca inalterado. En contraste, formatos como 3MF y GLB aproximan las superficies curvas mediante mallas de triángulos planos. Cuando se discretiza un cilindro circular de radio nominal en un polígono regular de N facetas, la desviación cordal máxima o sagita introduce una reducción radial calculada mediante la función trigonométrica correspondiente. A valores bajos de facetas, esta discrepancia dimensional puede impedir el ensamblaje adecuado de pasadores y bujes mecánicos. Al combinar scripts de modelado paramétrico en Python con verificadores deterministas que operan a nivel de bytes, los ingenieros independientes pueden certificar la equivalencia volumétrica entre representaciones continuas y discretas, garantizando tolerancias de holgura controladas sin depender de costosas plataformas comerciales.
Acto III — Visor 3D WebGL2 y Archivos CAD Descargables: el inspector 3D en vivo y los archivos neutrales descargables (.glb, .step, .3mf, .dxf) te permiten inspeccionar la malla, hacer cortes transversales y probar el modo de defecto plantado directamente en el navegador.
Actos IV y V — Kit para pequeños estudios, generador en Python y fuentes: descarga skills/toolchain/x02/make_x02.py, skills/toolchain/x02/verify_x02_toolchain_x02_parametric_step_3mf_roundtrip.py y SKILL.md. Cálculo educativo y exploratorio únicamente, no un diseño estructural firmado.
Act I — The Tooling Problem, Open Formats & Historical Turning Points (X.02)
Think of designing a precision metal bracket like writing a recipe. Instead of drawing fixed lines, you write down clear rules. You specify that the base is sixty millimeters long and the mounting boss sits centered on top. If you change one measurement, the entire part updates automatically. That is what parametric modeling means. But trouble starts when you share your file. Different manufacturing tools speak completely different languages. Some tools expect true curved surfaces, while others convert everything into flat triangular facets. When a round hole turns into flat facets, the opening shrinks slightly. Pins that should slide smoothly get stuck. Our free interactive tool solves this headache. Right in your browser, you can inspect an exact stepped pedestal in three dimensions. You can adjust the facet count and watch how the roundness changes in real time. In seconds, you can download verified files and Python scripts to audit your own parts.
From Cambridge B-Rep Kernels to Version-Controlled Python CAD & STEP AP242
In 1973 at the University of Cambridge, Ian Braid and Charles Lang asked how a computer could represent a machined metal part not as a hollow wireframe drawing, but as an unmistakable mathematical solid. Their answer—Boundary Representation (B-rep)—defined every solid by its exact analytical surfaces (planes, cylinders, cones, and NURBS patches) stitched together along topological edges and closed vertex loops.
For decades, B-rep kernels were locked behind expensive commercial licenses and incompatible binary files, while 3D printers relied on the 1987 STL format that stripped away units, topology, and curved surfaces. Between 2014 and 2026, three open revolutions converged: ISO 10303-242 (STEP AP242) unified neutral B-rep exchange, the 3MF Consortium created a unit-aware XML/ZIP mesh container, and open-source Python frameworks (CadQuery and build123d atop OpenCascade) turned parametric CAD into clean, Git-versioned Python scripts.
Introduces stereolithography STL files storing unindexed triangle coordinates for early additive layer slicing.
Classical Roadblock: STL omitted units (mm vs inch), vertex connectivity, and chordal error bounds.Unifies AP203 and AP214 into STEP AP242 Managed Model-Based 3D Engineering with explicit CLOSED_SHELL and PMI entities.
Classical Roadblock: Converting smooth STEP cylinders to meshes introduces chordal sagitta shrinkage.Replaces STL with a compact ZIP archive containing shared vertex tables, explicit millimeter units, and provenance metadata.
Classical Roadblock: Standard ZIP timestamps cause byte drift unless pinned deterministically.Certifies V = 43,200 mm³ parity across STEP AP242, 3MF, and GLB while bounding sagitta error h = R(1 − cos(π/N)) ≤ 0.0722 mm.
Classical Roadblock: Educational geometry only; CNC and additive tolerances require shop calibration.In the drafty computer laboratories of Cambridge University during the early 1970s, Ian Braid and Charles Lang wrestled with a quiet revolution on glowing cathode-ray screens. For generations, machinists and patternmakers had shaped turbine housings and gearbox casings using chalk markings, wooden templates, and steel calipers. Translating those curved, interlocking metal solids into computer memory seemed almost impossible. Wireframe sketches looked like ghostly spiderwebs that could not tell solid metal from empty air, while faceted polygon approximations stripped away the silky smooth curves of cylindrical bores. Alongside Hungarian researcher Timothy Varady in Budapest, Braid helped pioneer boundary representation, defining physical objects by their true geometric boundaries: smooth analytical planes, cylinders, and topological loops. Decades later, international standards teams codified that vision into the universal STEP AP242 standard, ensuring that precise geometric forms could travel freely between different software tools without losing their exact analytical curvature or dimensional truth.
For decades, mechanical and structural engineering remained tethered to expensive, closed-source CAD ecosystems that stored geometry in proprietary binary formats. Sharing a model between different modeling programs frequently resulted in translated geometry losing its parametric history, trimming curves failing to evaluate, and cylindrical surfaces collapsing into crude triangle approximations. The ratification of ISO 10303-21 STEP AP242 (DOI: DOI 10.1016/j.cad.2016.07.003 [Scite ↗]) — established an open, model-based engineering foundation for boundary representations. Instead of approximating curved shapes as flat polygons, STEP AP242 defines solids through explicit topological hierarchies: CARTESIAN_POINT, VERTEX_POINT, POLY_LOOP, and ADVANCED_FACE entities bounded within a CLOSED_SHELL. In Folio X.02, we demonstrate how open Python libraries such as build123d and CadQuery liberate engineers from commercial lock-in. A single version-controlled Python script generates our stepped structural clevis pedestal, producing an exact ISO STEP AP242 boundary representation alongside modern 3MF and binary GLB mesh files. Our byte-level verifier inspects all three formats simultaneously, confirming that analytical volumes match exactly while measuring the chordal tessellation error introduced whenever smooth analytic surfaces are converted into faceted meshes.
For an independent mechanical consultant or a small hardware team, file exchange discrepancies between software tools create costly friction. When a client sends a model exported through an unchecked conversion pipeline, subtle geometric corruptions can slip through unnoticed. A circular cylindrical bore faceted into thirty-two chords shrinks by more than seventy microns along its radius. If an assembler tries to slide a hardened steel locating dowel into that faceted opening, the pin seizes tight, halting physical assembly. Worse still, proprietary CAD vendors charge upwards of five thousand dollars per seat each year, locking solo practitioners out of professional revision-control and file-checking workflows. Open, script-driven CAD coupled with offline verification democratizes this entire engineering cycle. By parameterizing parts in readable Python code and executing byte-level audits across STEP, 3MF, and GLB exports, solo practitioners can mathematically prove geometric integrity, verify clearance limits, and collaborate seamlessly with machining facilities without costly proprietary software.
Independent Engineer's Zero-Cost Stack (build123d Apache-2.0 + FreeCAD 1.0 + OrcaSlicer): Author parametric parts in Python, export ISO 10303-21 STEP AP242 for CNC machine shops and 3MF for additive prototyping, and run verify_x02 to confirm zero volume drift between B-rep and mesh.
§ I.1 · Classical Obstruction, Open Standards & Historical Turning Points
Every enduring engineering and architectural tool begins as a human effort to turn tacit workshop craft into an inspectable, shareable language. Tracing the evolution of Toolbox X.02 (TOOL-ENG-1) from its earliest formulation in 1973–1980 (Ian Braid (Cambridge) and Charles Eastman pioneer boundary-representation (B-rep) and constructive solid geometry (CSG) kernels.) to its modern open-toolchain implementation in 2018–2026 (CadQuery 2 and build123d (Apache-2.0) bring algebraic Python scripting to OpenCascade, liberating solo engineers from proprietary CAD lock-in.) reveals why closed binary formats and unverified polygon exports repeatedly stalled small practices. Whenever a studio or fabricator is forced to exchange geometry without an explicit mathematical invariant—here governed by h_sagitta = R (1 − cos(π / N)) · V_step = V_flange + V_boss = 43,200 mm³ · ΔV / V < 10⁻⁶—silent topological defects such as boundary cracks, flipped face normals, or missing clearance tolerances corrupt the handoff between design and physical making.
Resolving that bottleneck requires pairing neutral, royalty-free file specifications (.glb, .step AP242, .3mf, .dxf) with a zero-dependency code generator (skills/toolchain/x02/make_x02.py) and a deterministic byte-level verifier (skills/toolchain/x02/verify_x02_toolchain_x02_parametric_step_3mf_roundtrip.py). Within the parametric state space governed by parameter P_1 (Circumferential Tessellation Facets N) spanning [8, 64] (reference N = 32); and parameter P_2 (Upper Boss Height H_boss) spanning [10, 30] (reference mm18), every exported file is treated not as an opaque drawing, but as a falsifiable mathematical certificate whose chronology of breakthroughs is summarized below:
| Year | Milestone & Verification Anchor |
|---|---|
| 1973–1980 | Ian Braid (Cambridge) and Charles Eastman pioneer boundary-representation (B-rep) and constructive solid geometry (CSG) kernels. |
| 1994–2014 | ISO 10303 (STEP) evolves from AP203/AP214 to AP242 Managed Model-Based 3D Engineering, unifying product manufacturing information (PMI) and exact B-rep topology. |
| 2015 | The 3MF Consortium releases the open XML/ZIP 3MF specification to replace unitless, non-manifold 1987 STL files. |
| 2018–2026 | CadQuery 2 and build123d (Apache-2.0) bring algebraic Python scripting to OpenCascade, liberating solo engineers from proprietary CAD lock-in. |
Code-first CAD turns a mechanical drawing from an opaque binary blob into a pure function: parameters go in, and verified STEP AP242, 3MF, and GLB artifacts come out with known chordal error bounds.
Act II — 4-Stage Geometry Synthesis & Byte-Level Verification Spine (X.02)
Everyday Analogy: Picture a master carpenter cutting a traditional Japanese watari-ago notched timber joint with a razor-sharp hand chisel. When shaping the interlocking tenon and mortise, every wooden face must be planed perfectly flat, meeting cleanly at right angles. If the carpenter pares away just a quarter-millimeter too much wood, the joint wobbles loosely and loses its grip. If the cut remains too tight, the beams refuse to slide together. A true solid CAD model demands that same exacting tactile fit: surfaces must meet with zero accidental gaps, preserving exact boundaries and precise physical clearances.
How the Two Sliders & Knockout Work in Plain Words: Before diving into technical formulas, take a look at the two controls and the counterfactual test in our interactive workbench. The first slider adjusts the Circumferential Tessellation Facets, labeled N. When converting a round hole or cylindrical boss into a 3D mesh, the computer approximates the circle with a polygon of N sides. At low values like eight or twelve, the circle looks noticeably chunky and shrinks inward, cutting into your clearance tolerance. As you slide N up toward thirty-two or sixty-four, the faceted shape hugs the true cylinder closely. The second slider changes the Upper Boss Height, growing or shrinking the top stepped pedestal. Watch how the exact mathematical volume recalculates instantly while keeping the base flange unchanged. Finally, try the Planted-Defect Knockout button. Clicking it intentionally drops the facet count or corrupts the schema, immediately breaching the tolerance threshold and triggering an audit failure that alerts the designer before any material is cut.
The 8-Sided Stop Sign Inside a Round Peg Hole
Draw a circle of radius R = 15 mm on graph paper—that is your exact STEP AP242 cylindrical pin hole. Now approximate that circle by drawing an 8-sided stop sign (N = 8 facets) whose corners touch the circle. Look at the flat straight edges of the stop sign: every flat chord cuts INSIDE the circle, shrinking the opening by the sagitta h = 15 × (1 − cos(180°/8)) = 1.14 mm!
If you try to slide a 30 mm steel dowel pin into that 8-sided faceted hole, the flat walls jam hard against the pin. Increase N to 32 facets in the workbench below and watch the chordal sagitta shrink by 16× down to 0.0722 mm—small enough for a clearance fit.
When an exact boundary representation is converted into a triangle mesh for visualization or additive fabrication, curved geometries undergo inevitable discretization error. Consider a circular cylindrical bore or boss of nominal radius R approximated by an inscribed regular polygon with N flat facets. The distance between the true circular arc and the center of each inscribed facet chord is governed by the chordal sagitta formula: h = R(1 − cos(π / N)). Using Taylor series expansion for large N, this geometric deviation scales inversely with the square of facet count: h ≈ R π² / (2 N²). For a cylindrical locating bore with radius R = 15 mm, setting N = 12 creates a chordal sagitta of 0.511 mm—an error massive enough to prevent mating pins from fitting. Doubling N to 24 reduces the sagitta to 0.128 mm, while N = 32 reaches 0.0722 mm, fitting within standard sliding tolerances. In contrast, planar stepped B-rep faces retain their exact analytic volume V = 43,200.0 mm³ identically across STEP AP242, 3MF, and GLB containers, illustrating the mathematical distinction between exact boundary geometry and faceted representations.
Live parameter coupling: N = 32
and 18 mm
under governing law h_sagitta = R (1 − cos(π / N)) · V_step = V_flange + V_boss = 43,200 mm³ · ΔV / V < 10⁻⁶.
Inside a boundary representation kernel, solid bodies are not stored as loose surface meshes, but as evaluated topological graphs anchored by precise geometric definitions. A closed solid shell is constructed through the Boolean union of oriented half-spaces, where each planar face is bounded by a directed closed wire known in ISO 10303-21 syntax as a POLY_LOOP. In our stepped clevis pedestal, the base flange (sixty by forty by twelve millimeters) merges with the upper boss (forty by twenty by eighteen millimeters) along a shared planar boundary at elevation z = 12 mm. Rather than retaining an internal dividing wall, the kernel executes half-space regularization, dissolving redundant internal faces and sewing outer faces into a single MANIFOLD_SOLID_BREP bounded by a CLOSED_SHELL entity. Each ADVANCED_FACE explicitly references its underlying surface equation, boundary orientation flag, and bounding vertex loop. This topological rigor guarantees that every edge is shared by exactly two adjacent faces with opposing parameter orientations, ensuring that analytical mass properties and volume integrals evaluate without approximation drift across neutral CAD exchange formats.
§ II.1 · Governing Geometric Law, Manifold Topology & Provenance Contract
h_sagitta = R (1 − cos(π / N)) · V_step = V_flange + V_boss = 43,200 mm³ · ΔV / V < 10⁻⁶
In Toolbox X.02 (Code-First Parametric Solids & Measured STEP AP242, 3MF, and GLB Round-Trip Fidelity), source of truth is code (skills/toolchain/x02/make_x02.py), never an unverified binary blob. Let the parametric family be governed by parameter P_1 (Circumferential Tessellation Facets N) spanning [8, 64] (reference N = 32); and parameter P_2 (Upper Boss Height H_boss) spanning [10, 30] (reference mm18). The generator emits deterministic, spec-compliant neutral files while the stdlib verifier skills/toolchain/x02/verify_x02_toolchain_x02_parametric_step_3mf_roundtrip.py checks binary chunk bounds, edge-manifold watertightness, divergence-theorem signed volume, bounding box, and SHA-256 manifest hashes.
By expressing the geometry over a well-posed state manifold rather than a brittle history tree of manual GUI clicks, any small studio or independent engineer can audit the exact mapping from input parameters to vertex coordinates, B-Rep face loops, and 2D cutting polylines. Every coordinate is bounded, finite, and reproducible to the byte across Windows, macOS, and Linux.
§ II.2 · Four-Stage Authoring & Verification Pipeline (Stages 1–4)
Rather than treating h_sagitta = R (1 − cos(π / N)) · V_step = V_flange + V_boss = 43,200 mm³ · ΔV / V < 10⁻⁶ as an opaque black box, we decompose the geometry generation and byte-level verification of Toolbox X.02 into four explicit, inspectable stages. True engineering and architectural tools do not ask for blind trust; they invite inspection at every step:
Stage 1 Geometry & Verification Pipeline
Formal Stage Mechanism: Stage 1: Algebraic Python Solid Definition — Parameterize the stepped flange (60 × 40 × 12 mm) and boss (40 × 20 × 18 mm) as a single closed 2-manifold boundary with V_exact = 28,800 + 14,400 = 43,200.0 mm³.
Byte-Level Invariant Closure: At Stage 1 of Toolbox X.02 (TOOL-ENG-1), the generator skills/toolchain/x02/make_x02.py and stdlib verifier skills/toolchain/x02/verify_x02_toolchain_x02_parametric_step_3mf_roundtrip.py enforce the governing geometric law h_sagitta = R (1 − cos(π / N)) · V_step = V_flange + V_boss = 43,200 mm³ · ΔV / V < 10⁻⁶ along with the companion invariant: Cross-Format Topological Identity: The vertex count (V = 16), face count (F = 28), and Euler characteristic (χ = 2) must match identically across x02_artifact.step, x02_artifact.3mf, and x02_artifact.glb.
Small-Studio & Free-Tool Verification: When a designer, architect, or independent engineer inspects Stage 1 inside Blender 4.2+, FreeCAD 1.0, OrcaSlicer / PrusaSlicer, or QCAD / LibreCAD—or executes python skills/toolchain/x02/verify_x02_toolchain_x02_parametric_step_3mf_roundtrip.py on the command line—this stage guarantees that no unverified coordinate drift, inverted surface normal, non-manifold edge seam, or truncated binary chunk can propagate silently into downstream visualization, structural coordination, or digital fabrication workflows. Every output file in src/research/toolchain/x02/files/ is locked by a deterministic SHA-256 hash in manifest.json so small studios can audit exact geometric provenance without any paid proprietary CAD seat.
Stage 2 Geometry & Verification Pipeline
Formal Stage Mechanism: Stage 2: ISO 10303-21 STEP AP242 B-Rep Serialization — Emit explicit CARTESIAN_POINT, VERTEX_POINT, POLY_LOOP, ADVANCED_FACE, CLOSED_SHELL, and MANIFOLD_SOLID_BREP entities with millimeter length units (#7) and 10⁻⁵ mm uncertainty (#6).
Byte-Level Invariant Closure: At Stage 2 of Toolbox X.02 (TOOL-ENG-1), the generator skills/toolchain/x02/make_x02.py and stdlib verifier skills/toolchain/x02/verify_x02_toolchain_x02_parametric_step_3mf_roundtrip.py enforce the governing geometric law h_sagitta = R (1 − cos(π / N)) · V_step = V_flange + V_boss = 43,200 mm³ · ΔV / V < 10⁻⁶ along with the companion invariant: Chordal Sagitta Tolerance Bound: For any cylindrical feature of radius R discretized into N segments, the radial deviation is strictly h = R(1 − cos(π/N)); keeping h ≤ 0.10 mm at R = 15 mm requires N ≥ 28.
Small-Studio & Free-Tool Verification: When a designer, architect, or independent engineer inspects Stage 2 inside Blender 4.2+, FreeCAD 1.0, OrcaSlicer / PrusaSlicer, or QCAD / LibreCAD—or executes python skills/toolchain/x02/verify_x02_toolchain_x02_parametric_step_3mf_roundtrip.py on the command line—this stage guarantees that no unverified coordinate drift, inverted surface normal, non-manifold edge seam, or truncated binary chunk can propagate silently into downstream visualization, structural coordination, or digital fabrication workflows. Every output file in src/research/toolchain/x02/files/ is locked by a deterministic SHA-256 hash in manifest.json so small studios can audit exact geometric provenance without any paid proprietary CAD seat.
Stage 3 Geometry & Verification Pipeline
Formal Stage Mechanism: Stage 3: Deterministic 3MF ZIP & GLB Binary Packaging — Package the identical 16-vertex, 28-triangle topology into 3MF Core 1.2.3 XML (fixed ZIP timestamp) and Khronos GLB binary buffers with embedded provenance metadata.
Byte-Level Invariant Closure: At Stage 3 of Toolbox X.02 (TOOL-ENG-1), the generator skills/toolchain/x02/make_x02.py and stdlib verifier skills/toolchain/x02/verify_x02_toolchain_x02_parametric_step_3mf_roundtrip.py enforce the governing geometric law h_sagitta = R (1 − cos(π / N)) · V_step = V_flange + V_boss = 43,200 mm³ · ΔV / V < 10⁻⁶ along with the companion invariant: Provenance & Planted-Defect Gate: Every exported file must embed the generator script SHA-256 and boundary statement, and verify_x02 must reject any flipped-normal or corrupted-schema mutation.
Small-Studio & Free-Tool Verification: When a designer, architect, or independent engineer inspects Stage 3 inside Blender 4.2+, FreeCAD 1.0, OrcaSlicer / PrusaSlicer, or QCAD / LibreCAD—or executes python skills/toolchain/x02/verify_x02_toolchain_x02_parametric_step_3mf_roundtrip.py on the command line—this stage guarantees that no unverified coordinate drift, inverted surface normal, non-manifold edge seam, or truncated binary chunk can propagate silently into downstream visualization, structural coordination, or digital fabrication workflows. Every output file in src/research/toolchain/x02/files/ is locked by a deterministic SHA-256 hash in manifest.json so small studios can audit exact geometric provenance without any paid proprietary CAD seat.
Stage 4 Geometry & Verification Pipeline
Formal Stage Mechanism: Stage 4: Cross-Format Round-Trip & Chordal Deviation Audit — Verify that STEP, 3MF, and GLB agree on vertex/face counts and analytical volume, and bound circular bore discretization error via h = R(1 − cos(π/N)).
Byte-Level Invariant Closure: At Stage 4 of Toolbox X.02 (TOOL-ENG-1), the generator skills/toolchain/x02/make_x02.py and stdlib verifier skills/toolchain/x02/verify_x02_toolchain_x02_parametric_step_3mf_roundtrip.py enforce the governing geometric law h_sagitta = R (1 − cos(π / N)) · V_step = V_flange + V_boss = 43,200 mm³ · ΔV / V < 10⁻⁶ along with the companion invariant: Cross-Format Topological Identity: The vertex count (V = 16), face count (F = 28), and Euler characteristic (χ = 2) must match identically across x02_artifact.step, x02_artifact.3mf, and x02_artifact.glb.
Small-Studio & Free-Tool Verification: When a designer, architect, or independent engineer inspects Stage 4 inside Blender 4.2+, FreeCAD 1.0, OrcaSlicer / PrusaSlicer, or QCAD / LibreCAD—or executes python skills/toolchain/x02/verify_x02_toolchain_x02_parametric_step_3mf_roundtrip.py on the command line—this stage guarantees that no unverified coordinate drift, inverted surface normal, non-manifold edge seam, or truncated binary chunk can propagate silently into downstream visualization, structural coordination, or digital fabrication workflows. Every output file in src/research/toolchain/x02/files/ is locked by a deterministic SHA-256 hash in manifest.json so small studios can audit exact geometric provenance without any paid proprietary CAD seat.
§ II.3 · Deterministic Byte-Level Invariants & Planted-Defect Gate
Every neutral file and numerical readout in Toolbox X.02 is governed by three machine-checked falsification invariants embedded in toolchain-x02-parametric-step-3mf-roundtrip-SKILL.md and executed offline by skills/toolchain/x02/verify_x02_toolchain_x02_parametric_step_3mf_roundtrip.py:
-
Invariant X.02-R1: Cross-Format Topological Identity: The vertex count (V = 16), face count (F = 28), and Euler characteristic (χ = 2) must match identically across x02_artifact.step, x02_artifact.3mf, and x02_artifact.glb.
Checked deterministically offline by
skills/toolchain/x02/verify_x02_toolchain_x02_parametric_step_3mf_roundtrip.pyagainst committed binary artifacts insrc/research/toolchain/x02/files/. -
Invariant X.02-R2: Chordal Sagitta Tolerance Bound: For any cylindrical feature of radius R discretized into N segments, the radial deviation is strictly h = R(1 − cos(π/N)); keeping h ≤ 0.10 mm at R = 15 mm requires N ≥ 28.
Checked deterministically offline by
skills/toolchain/x02/verify_x02_toolchain_x02_parametric_step_3mf_roundtrip.pyagainst committed binary artifacts insrc/research/toolchain/x02/files/. -
Invariant X.02-R3: Provenance & Planted-Defect Gate: Every exported file must embed the generator script SHA-256 and boundary statement, and verify_x02 must reject any flipped-normal or corrupted-schema mutation.
Checked deterministically offline by
skills/toolchain/x02/verify_x02_toolchain_x02_parametric_step_3mf_roundtrip.pyagainst committed binary artifacts insrc/research/toolchain/x02/files/.
Together, Invariants X.02-R1 through X.02-R3 close the verification loop between parametric authoring and neutral CAD exchange. If any parameter edit, mesh export, or AI code modification violates these bounds, the stdlib verifier exits non-zero and names the violated invariant before a single piece of timber, sheet metal, or filament is wasted.
§ II.4 · Planted-Defect Knockout & Failure-Mode Anatomy
Primary Structural Note (SIDENOTE X.02-A · WHY STL SHOULD RETIRE IN FAVOR OF 3MF & STEP): The 1987 STL format stores raw triangle soup with no units, no vertex sharing, no material metadata, and no license/provenance block. 3MF and STEP AP242 preserve explicit units (mm) and shared vertex indices.
Secondary Verification Note (SIDENOTE X.02-B · APACHE-2.0 BUILD123D FOR SOLO ENGINEERS): Unlike GPL-licensed script CAD tools whose copyleft terms can complicate client deliverables, build123d and CadQuery are licensed under Apache-2.0, making them clean for independent consulting engineers.
Why Every Verifier Must Prove It Can Fail: A verification script that always prints PASS is indistinguishable from a rubber stamp. That is why skills/toolchain/x02/verify_x02_toolchain_x02_parametric_step_3mf_roundtrip.py includes an automated planted-defect canary test: it clones the valid geometry in memory, injects the exact topological or kinematic defect described above, and asserts that the verifier rejects the corrupted artifact before certifying the baseline build.
Act III — Live WebGL2 3D Inspector, Downloadable CAD Files & Interactive Workbench
§ III.1 · Three Guided Experiments to Try
- Inspect the stepped clevis pedestal in the live WebGL2 viewer below; toggle Wireframe to see the seamless 16-vertex, 28-triangle shoulder transition at z = 12 mm.
- Scrub Circumferential Facets N down from 32 to 8 (or click [⚡ Counterfactual Axiom-Knockout]) to watch the chordal sagitta error h = R(1 − cos(π/N)) breach the 0.10 mm pin-clearance budget.
- Download x02_artifact.step, x02_artifact.3mf, and x02_artifact.glb to test import fidelity in FreeCAD, Blender, or your slicer.
§ III.2 · Downloadable Neutral 3D CAD & Fabrication Files + Provenance Manifest
Every file below is generated deterministically by skills/toolchain/x02/make_x02.py and verified byte-for-byte against manifest.json. Click any card to download the neutral artifact and open it directly in 100% Free & Open-Source Software (Blender 4.2+, FreeCAD 1.0, Bonsai BIM, OrcaSlicer, PrusaSlicer, QCAD Community Edition, LibreCAD, Inkscape):
Certified Geometry Regime (X.02) — All 4 stage invariants satisfied (skills/toolchain/x02/verify_x02_toolchain_x02_parametric_step_3mf_roundtrip.py).
Drag either slider above to scale the live 3D WebGL2 mesh in real time, then click any Export button to synthesize a custom-sized .glb, .step, or .dxf file directly in your browser.
§ III.3 · Worked Byte-Level Verification & Reference Telemetry Table
To verify the interactive 3D Inspector and Workbench above by hand or inside an automated studio QA pipeline, we evaluate h_sagitta = R (1 − cos(π / N)) · V_step = V_flange + V_boss = 43,200 mm³ · ΔV / V < 10⁻⁶ at the canonical baseline operating point (parameter P_1 (Circumferential Tessellation Facets N) spanning [8, 64] (reference N = 32); and parameter P_2 (Upper Boss Height H_boss) spanning [10, 30] (reference mm18)) on the committed artifacts in src/research/toolchain/x02/files/. The resulting geometric and topological telemetry is certified as follows: Exact B-Rep / Mesh Volume certifies 43,200.0 mm³ (ΔV = 0.00%); Chordal Sagitta Error h(N) certifies 0.0722 mm (at R=15, N=32); STEP AP242 Schema Audit certifies CLOSED_SHELL · 28 FACES PASS; 3MF & GLB Parity Verdict certifies PASS · SHA-256 VERIFIED.
| Geometry / Format Check | Certified Artifact Value | Stdlib Verifier |
|---|---|---|
| Exact B-Rep / Mesh Volume | 43,200.0 mm³ (ΔV = 0.00%) | skills/toolchain/x02/verify_x02_toolchain_x02_parametric_step_3mf_roundtrip.py · PASS |
| Chordal Sagitta Error h(N) | 0.0722 mm (at R=15, N=32) | skills/toolchain/x02/verify_x02_toolchain_x02_parametric_step_3mf_roundtrip.py · PASS |
| STEP AP242 Schema Audit | CLOSED_SHELL · 28 FACES PASS | skills/toolchain/x02/verify_x02_toolchain_x02_parametric_step_3mf_roundtrip.py · PASS |
| 3MF & GLB Parity Verdict | PASS · SHA-256 VERIFIED | skills/toolchain/x02/verify_x02_toolchain_x02_parametric_step_3mf_roundtrip.py · PASS |
Running python skills/toolchain/x02/make_x02.py followed by python skills/toolchain/x02/verify_x02_toolchain_x02_parametric_step_3mf_roundtrip.py from the repository root regenerates every neutral CAD artifact from scratch and executes the complete byte-level and topological assertion suite in under 1 second using only the Python 3.11+ standard library.
Act IV — Small-Studio Empowerment, T0–T3 Truth Boundary & Bridged Folios (X.02)
In modern parametric CAD engineering, artificial intelligence tools can generate complex solid modeling scripts in Python upon request. However, large language models frequently invent nonexistent API methods, misplace Boolean coordinate offsets, or hallucinate topological connections that produce invalid non-manifold geometry or unstitched boundary shells. Folio X.02 introduces a disciplined co-design workflow that prevents these failure modes. The human engineer defines the mechanical interface requirements, dimensional tolerances, and aesthetic proportions. The AI coding agent implements the parametric recipe in Python using build123d or CadQuery syntax, adhering to portable rules documented in our companion SKILL.md guide. Before any generated STEP AP242, 3MF, or GLB file is accepted into production, verify_x02 executes a fully automated, offline verification pass. The verifier parses the STEP entity graph, audits XML triangle indices in 3MF archives, validates binary GLB buffers, and checks the chordal sagitta error against specified pin-clearance tolerances. This mathematical gatekeeper ensures that the generative speed of artificial intelligence is harnessed safely, leaving human engineers in complete command of physical intent and functional geometry.
The boundary-representation and round-trip verification methodologies developed in Folio X.02 form a vital connective bridge across Informatico.ai. In Civil and Structural Engineering, modeling elastomeric bearings for seismic base isolation (Folio E.06) demands precise parametric cylindrical boss geometry to verify lead-core confinement clearances under cyclic lateral shear displacement. In Advanced Manufacturing, toolpath synthesis for 3D concrete printing (Folio E.10) requires exact STEP B-rep solids to compute thixotropic bead slicing and avoid nozzle trajectory deviations. In Computational Mechanics, SIMP topology optimization (Folio E.11) outputs organic density distributions that must be reconstructed into smooth boundary-representation solids for CNC milling operations. In Mathematics, neuro-symbolic geometry synthesis (Folio M.05) bridges algebraic constraint systems with Euclidean constructive geometry, matching the dual representation of parametric code and evaluated B-rep entities. By mastering code-first CAD modeling in open Python, solo engineers acquire a unified computational toolchain capable of expressing complex physical concepts across civil, mechanical, and mathematical disciplines without proprietary constraints or commercial software barriers.
§ IV.1 · Certified Geometry in Repository (T0/T1) vs. Educational Boundary & Edition 2 Horizon (T2/T3)
Deterministic Generator, Neutral 3D Files & Byte-Level Verifier
- Deterministic generator (skills/toolchain/x02/make_x02.py) emitting byte-reproducible STEP AP242 (.step), 3MF Core 1.2.3 (.3mf), and glTF 2.0 (.glb) solids.
- Stdlib cross-format verifier (verify_x02_toolchain_x02_parametric_step_3mf_roundtrip.py) checking ISO-10303-21 AP242 B-rep entities, 3MF ZIP XML structure, GLB manifoldness, and V = 43,200.0 mm³.
- Explicit Round-Trip Survival Matrix documenting what survives across STEP AP242, 3MF, and GLB (units, topology, colors, provenance).
Geometry vs. Physical Fitness & Future Extensions
- Educational Geometry Boundary: Every file on this page is an educational artifact; nothing is certified as load-bearing, printable, safe, or fit for any purpose.
- Future Edition 2 Extensions: Analytic NURBS trimmed-surface STEP AP242 export via build123d/OCCT C++ bindings and automated PMI semantic tolerance extraction.
§ IV.2 · Bridged Pillar Folios & Permissive Open-Source Ecosystem
Every tool in Tech & Tools · The Toolbox is designed as an operational bridge across the four core pillars of Informatico.ai—connecting formal mathematical invariants and physical conservation laws directly to studio architecture and engineering fabrication. Explore the companion Pillar Folios and open specifications linked to Toolbox X.02 below:
All active runtime components for Toolbox X.02 are audited in skills/toolchain/licenses.json to ensure permissive licensing (MIT, Apache-2.0, MPL-2.0, LGPL-2.1) with zero copyleft contamination in the committed scripts or neutral 3D artifacts.
Act V — Reproducible Generator (`make_x02.py`), Stdlib Verifier & Portable SKILL.md
§ V.1 · Source-of-Truth Generator (skills/toolchain/x02/make_x02.py) & Verifier (skills/toolchain/x02/verify_x02_toolchain_x02_parametric_step_3mf_roundtrip.py)
Below are the complete, unabridged Python 3 scripts that generate and verify Toolbox X.02 (TOOL-ENG-1). Because code is the primary source of truth, you can inspect every equation, vertex buffer, STEP entity, and planted-defect assertion right on this page, or run them locally on your own machine:
Inspect generator script: skills/toolchain/x02/make_x02.py
#!/usr/bin/env python3
"""Deterministic Geometry Generator for Toolbox X.02 (TOOL-ENG-1): Code-First Parametric Solids & Measured STEP AP242, 3MF, and GLB Round-Trip Fidelity.
Canonical URL: https://informatico.ai/toolchain/x02/
Boundary: Every file on this page is an educational artifact generated by the script beside it. Nothing here is certified as load-bearing, printable, safe, or fit for any purpose. The verifier checks geometry, not fitness.
"""
from __future__ import annotations
import json
import sys
from pathlib import Path
ROOT = Path(__file__).resolve().parents[3]
sys.path.insert(0, str(ROOT / "tools"))
from toolchain_uga import generate_folio_assets # noqa: E402
def main() -> int:
files_dir = ROOT / "src" / "research" / "toolchain" / "x02" / "files"
manifest = generate_folio_assets(
slug="x02",
folio_id="X.02",
title="Code-First Parametric Solids & Measured STEP AP242, 3MF, and GLB Round-Trip Fidelity",
script_path=Path(__file__).resolve(),
files_dir=files_dir,
formats=('glb', 'step', '3mf'),
)
print(json.dumps({"status": "GENERATED", "folio_id": "X.02", "files": [f["filename"] for f in manifest["files"]]}, indent=2))
return 0
if __name__ == "__main__":
raise SystemExit(main())
Inspect byte-level verifier script: skills/toolchain/x02/verify_x02_toolchain_x02_parametric_step_3mf_roundtrip.py
#!/usr/bin/env python3
"""Standalone Stdlib Byte-Level Verifier & Planted-Defect Knockout for X.02 (TOOL-ENG-1): Code-First Parametric Solids & Measured STEP AP242, 3MF, and GLB Round-Trip Fidelity.
Canonical URL: https://informatico.ai/toolchain/x02/
Governing Law: h_sagitta = R (1 − cos(π / N)) · V_step = V_flange + V_boss = 43,200 mm³ · ΔV / V < 10⁻⁶
Boundary: Every file on this page is an educational artifact generated by the script beside it. Nothing here is certified as load-bearing, printable, safe, or fit for any purpose. The verifier checks geometry, not fitness.
"""
from __future__ import annotations
import json
import sys
from pathlib import Path
ROOT = Path(__file__).resolve().parents[3]
sys.path.insert(0, str(ROOT / "tools"))
from toolchain_uga import verify_folio_assets # noqa: E402
def main(argv: list[str] | None = None) -> int:
args = argv if argv is not None else sys.argv[1:]
planted_only = "--planted-defect" in args
files_dir = ROOT / "src" / "research" / "toolchain" / "x02" / "files"
make_script = Path(__file__).resolve().parent / "make_x02.py"
try:
result = verify_folio_assets(
slug="x02",
files_dir=files_dir,
script_path=make_script,
planted_defect_only=planted_only,
)
except AssertionError as exc:
print(json.dumps({"status": "FAIL", "folio_id": "X.02", "error": str(exc)}), file=sys.stderr)
return 1
print(json.dumps(result, indent=2, ensure_ascii=False))
return 0
if __name__ == "__main__":
raise SystemExit(main())
Equip your studio or AI coding agent with the tool-neutral workflow, open-format export rules, and byte-level verifier gate for Toolbox X.02 (TOOL-ENG-1): download toolchain-x02-parametric-step-3mf-roundtrip-SKILL.md or verify_x02_toolchain_x02_parametric_step_3mf_roundtrip.py.
Show the SKILL.md source
---
name: toolchain-x02-parametric-step-3mf-roundtrip
description: Deterministic 3D CAD generator, byte-level stdlib verifier, and small-studio open-tool kit for X.02 (TOOL-ENG-1) - Code-First Parametric Solids & Measured STEP AP242, 3MF, and GLB Round-Trip Fidelity.
license: MIT
compatibility: Claude Code, GitHub Copilot, Cursor, Google Antigravity, OpenAI Codex, Ollama (Qwen 3.8 / Gemma 4)
---
# Code-First Parametric Solids & Measured STEP AP242, 3MF, and GLB Round-Trip Fidelity (X.02 · TOOL-ENG-1)
> **Fixed Educational Boundary (`EN`):** Every file on this page is an educational artifact generated by the script beside it. Nothing here is certified as load-bearing, printable, safe, or fit for any purpose. The verifier checks geometry, not fitness.
>
> **Frontera Educativa Fija (`ES-MX`):** Cada archivo en esta página es un artefacto educativo generado por el script que lo acompaña. Nada aquí está certificado como estructural, imprimible, seguro ni apto para ningún propósito. El verificador comprueba la geometría, no la aptitud.
## 1. Governing Law, Open Standard & Canonical URL
- **Toolbox Folio**: `X.02` (`TOOL-ENG-1` · `REV 1.2`, Redated `2026-11-09`)
- **Canonical URL**: https://informatico.ai/toolchain/x02/ (Spanish: https://informatico.ai/es-mx/toolchain/x02/)
- **Track Badge**: TOOLBOX X.02 · BUILD123D / CADQUERY KERNEL → STEP AP242 · 3MF · GLB AUDIT
- **Reference Standard / Repository**: [gumyr/build123d (Apache-2.0 Python BREP CAD)](https://github.com/gumyr/build123d)
- **Governing Equation**: `h_sagitta = R (1 − cos(π / N)) · V_step = V_flange + V_boss = 43,200 mm³ · ΔV / V < 10⁻⁶`
- **Deterministic Generator**: `python skills/toolchain/x02/make_x02.py`
- **Byte-Level Stdlib Verifier**: `python skills/toolchain/x02/verify_x02_toolchain_x02_parametric_step_3mf_roundtrip.py`
- **Planted-Defect Knockout Check**: `python skills/toolchain/x02/verify_x02_toolchain_x02_parametric_step_3mf_roundtrip.py --planted-defect` (must exit non-zero)
- **Primary Scite Anchor DOI**: https://doi.org/10.1016/j.cad.2021.103018 (Scite Report: https://scite.ai/reports/10.1016/j.cad.2021.103018)
## 2. Small-Studio & Independent-Practitioner Open-Tool Workflow
- **EN**: Independent Engineer's Zero-Cost Stack (build123d Apache-2.0 + FreeCAD 1.0 + OrcaSlicer): Author parametric parts in Python, export ISO 10303-21 STEP AP242 for CNC machine shops and 3MF for additive prototyping, and run verify_x02 to confirm zero volume drift between B-rep and mesh.
- **ES-MX**: Stack sin Costo para Ingenieros Independientes (build123d Apache-2.0 + FreeCAD 1.0 + OrcaSlicer): Modela piezas paramétricas en Python, exporta STEP AP242 (ISO 10303-21) para maquinado CNC y 3MF para manufactura aditiva, y ejecuta verify_x02 para certificar cero deriva volumétrica.
## 3. Tool-Neutral Agent Packet & Security Boundary (`security-runtime`)
1. **Generate Deterministic 3D Artifacts Locally**: Run `python skills/toolchain/x02/make_x02.py` inside the repository working directory to emit `src/research/toolchain/x02/files/x02_artifact.*` and `manifest.json`.
2. **Run Byte-Level Verifier as Acceptance Gate**: Run `python skills/toolchain/x02/verify_x02_toolchain_x02_parametric_step_3mf_roundtrip.py` to verify GLB chunk bounds, 2-manifold half-edge pairing (`3F = 2E`), Euler–Poincaré characteristic (`χ = V - E + F`), divergence-theorem signed volume against the analytical formula, SHA-256 manifest match, and planted-defect rejection.
3. **Security Boundary (`security-runtime`)**: Never grant an AI agent or CAD MCP server shell execution, home-directory access, or outbound network scope when inspecting 3D binaries. Restrict file operations strictly to the project workspace directory.
## 4. Deterministic Verification & Falsification Rules
1. Cross-Format Topological Identity: The vertex count (V = 16), face count (F = 28), and Euler characteristic (χ = 2) must match identically across x02_artifact.step, x02_artifact.3mf, and x02_artifact.glb.
2. Chordal Sagitta Tolerance Bound: For any cylindrical feature of radius R discretized into N segments, the radial deviation is strictly h = R(1 − cos(π/N)); keeping h ≤ 0.10 mm at R = 15 mm requires N ≥ 28.
3. Provenance & Planted-Defect Gate: Every exported file must embed the generator script SHA-256 and boundary statement, and verify_x02 must reject any flipped-normal or corrupted-schema mutation.
## 5. 4-Stage Geometry & Verification Pipeline
- Stage 1: Algebraic Python Solid Definition — Parameterize the stepped flange (60 × 40 × 12 mm) and boss (40 × 20 × 18 mm) as a single closed 2-manifold boundary with V_exact = 28,800 + 14,400 = 43,200.0 mm³.
- Stage 2: ISO 10303-21 STEP AP242 B-Rep Serialization — Emit explicit CARTESIAN_POINT, VERTEX_POINT, POLY_LOOP, ADVANCED_FACE, CLOSED_SHELL, and MANIFOLD_SOLID_BREP entities with millimeter length units (#7) and 10⁻⁵ mm uncertainty (#6).
- Stage 3: Deterministic 3MF ZIP & GLB Binary Packaging — Package the identical 16-vertex, 28-triangle topology into 3MF Core 1.2.3 XML (fixed ZIP timestamp) and Khronos GLB binary buffers with embedded provenance metadata.
- Stage 4: Cross-Format Round-Trip & Chordal Deviation Audit — Verify that STEP, 3MF, and GLB agree on vertex/face counts and analytical volume, and bound circular bore discretization error via h = R(1 − cos(π/N)).
## 6. Certified Reference Telemetry (T0/T1)
- **Exact B-Rep / Mesh Volume**: `43,200.0 mm³ (ΔV = 0.00%)`
- **Chordal Sagitta Error h(N)**: `0.0722 mm (at R=15, N=32)`
- **STEP AP242 Schema Audit**: `CLOSED_SHELL · 28 FACES PASS`
- **3MF & GLB Parity Verdict**: `PASS · SHA-256 VERIFIED`
## 7. Exact T0–T3 Truth Boundary
### Certified Geometry in Repository Today (T0/T1)
- Deterministic generator (skills/toolchain/x02/make_x02.py) emitting byte-reproducible STEP AP242 (.step), 3MF Core 1.2.3 (.3mf), and glTF 2.0 (.glb) solids.
- Stdlib cross-format verifier (verify_x02_toolchain_x02_parametric_step_3mf_roundtrip.py) checking ISO-10303-21 AP242 B-rep entities, 3MF ZIP XML structure, GLB manifoldness, and V = 43,200.0 mm³.
- Explicit Round-Trip Survival Matrix documenting what survives across STEP AP242, 3MF, and GLB (units, topology, colors, provenance).
### Educational Boundary & Edition 2 Horizon (T2/T3)
- Educational Geometry Boundary: Every file on this page is an educational artifact; nothing is certified as load-bearing, printable, safe, or fit for any purpose.
- Future Edition 2 Extensions: Analytic NURBS trimmed-surface STEP AP242 export via build123d/OCCT C++ bindings and automated PMI semantic tolerance extraction.
#!/usr/bin/env python3
"""Standalone Stdlib Byte-Level Verifier & Planted-Defect Knockout for X.02 (TOOL-ENG-1): Code-First Parametric Solids & Measured STEP AP242, 3MF, and GLB Round-Trip Fidelity.
Canonical URL: https://informatico.ai/toolchain/x02/
Governing Law: h_sagitta = R (1 − cos(π / N)) · V_step = V_flange + V_boss = 43,200 mm³ · ΔV / V < 10⁻⁶
Boundary: Every file on this page is an educational artifact generated by the script beside it. Nothing here is certified as load-bearing, printable, safe, or fit for any purpose. The verifier checks geometry, not fitness.
"""
from __future__ import annotations
import json
import sys
from pathlib import Path
ROOT = Path(__file__).resolve().parents[3]
sys.path.insert(0, str(ROOT / "tools"))
from toolchain_uga import verify_folio_assets # noqa: E402
def main(argv: list[str] | None = None) -> int:
args = argv if argv is not None else sys.argv[1:]
planted_only = "--planted-defect" in args
files_dir = ROOT / "src" / "research" / "toolchain" / "x02" / "files"
make_script = Path(__file__).resolve().parent / "make_x02.py"
try:
result = verify_folio_assets(
slug="x02",
files_dir=files_dir,
script_path=make_script,
planted_defect_only=planted_only,
)
except AssertionError as exc:
print(json.dumps({"status": "FAIL", "folio_id": "X.02", "error": str(exc)}), file=sys.stderr)
return 1
print(json.dumps(result, indent=2, ensure_ascii=False))
return 0
if __name__ == "__main__":
raise SystemExit(main())
§ V.2 · Primary Peer-Reviewed Sources & Open Specification Anchors
- ISO 10303-242:2022. Industrial automation systems and integration — Product data representation and exchange — Part 242: Application protocol: Managed model-based 3D engineering.
- Feeney, A. B., Frechette, S. P., & Srinivasan, V. (2015). A portrait of an ISO STEP tolerancing standard as an enabler of smart manufacturing systems. Journal of Computing and Information Science in Engineering, 15(2), 021001. DOI DOI 10.1115/1.4029050 [Scite ↗]
- 3MF Consortium. (2023). 3D Manufacturing Format Core Specification & Reference Guide v1.2.3. github.com/3MFConsortium/spec_core
- build123d Contributors. (2023–2026). build123d: A python CAD programming library (Apache-2.0). github.com/gumyr/build123d