Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add beiming183-cloud/AutoCAD-MCP --skill industrial-product-design-gbtgit clone --depth 1 https://github.com/beiming183-cloud/AutoCAD-MCPWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/beiming183-cloud/autocad-mcp/industrial-product-design-gbt)<a href="https://agentmods.dev/skills/beiming183-cloud/autocad-mcp/industrial-product-design-gbt"><img src="https://agentmods.dev/badge/skills/beiming183-cloud/autocad-mcp/industrial-product-design-gbt/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/beiming183-cloud/autocad-mcp/industrial-product-design-gbt"><img src="https://agentmods.dev/badge/skills/beiming183-cloud/autocad-mcp/industrial-product-design-gbt.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00136 | $0.04871 |
| Opus 5 | $0.00068 | $0.02436 |
| Sonnet 5 | $0.00027 | $0.00974 |
| Haiku 4.5 | $0.00014 | $0.00487 |
Grade A, and why
industrial-product-design-gbt scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 12d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 232 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Industrial Product Design with GB/T Handoff
Treat product design as the primary task and mechanical drafting as the engineering expression of a sufficiently resolved design. Decide who the product serves, what it must do, how it is used, why it has its form, and how it will be made and verified before detailed CAD and drawings.
Core Principles
- Converge function, architecture, interaction, form, CMF, manufacturing, service, and safety together. Do not apply styling as a late shell over an unresolved package.
- Separate observation, inference, assumption, and confirmed requirement. Mark unknown inputs
TBDand checks that did not runNOT_EVALUATED. - Create meaningful alternatives. Color or fillet variants of the first idea do not count as concept exploration.
- Express aesthetics as testable form language: proportion, silhouette, mass hierarchy, geometry vocabulary, line and surface relationships, transitions, boundaries, detail density, CMF, and 360-degree coherence.
- Keep editable parametric native CAD as engineering truth. Renders, meshes, images, plots, and exchange files are derived evidence.
- Never infer manufacturing-critical dimensions, material performance, tolerances, safety distances, or compliance from appearance alone.
- Lock the product's immutable design DNA before optimization. Do not solve usability, packaging, cost, or engineering pressure by silently deleting the defining innovation.
- Label every output by maturity. A concept is not a manufacturable design, and automated checks do not grant release approval.
- Never treat body count, entity count, dimension audit, geometry DRC, or save success as proof that proportion, form language, interface integration, motion logic, surface quality, or product realism is acceptable.
Portability Contract
- Discover the available CAD application, automation bridge, renderers, parsers, and file formats before execution. Map capability-level steps to tools that actually exist.
- Never invent a tool call or verification result. After mutable CAD operations, read back actual entities, parameters, associations, hierarchy, and geometry.
- Preserve the user's source and unrelated geometry. Use a traceable working copy for major revisions; do not overwrite, remove, or replace user files without authority.
- Keep deliverables and evidence in standard CAD, exchange, image, and report artifacts that another tool or reviewer can reopen.
What ships with it
43 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- .gitignore 57 B
- agents/openai.yaml 335 B
- assets/camera-presets.json 1.4 KB
- assets/local-learning-settings.example.json 606 B
- assets/project-feedback-template.json 1.7 KB
- references/aesthetic-profile.md 4.5 KB
- references/autocad-mcp-workflow.md 21 KB
- references/backend-routing.md 4.2 KB
- references/cad-3d-modeling.md 9.3 KB
- references/cad-workflows.md 13 KB
- references/complex-assembly-drafting.md 12 KB
- references/consumer-product-concept.md 16 KB
- references/design-brief-and-gates.md 7.3 KB
- references/drc-review.md 21 KB
- references/engineering-analysis.md 7.5 KB
- references/failure-learning.md 2.5 KB
- references/form-language-and-aesthetics.md 4.6 KB
- references/gbt-drafting.md 15 KB
- references/gps-inspection.md 7.4 KB
- references/human-factors-and-usability.md 4.1 KB
- references/industrial-design-3d-quality.md 14 KB
- references/integrated-design-review.md 5.2 KB
- references/local-cad-reference-library.md 6.2 KB
- references/local-reference-library.example.json 191 B
- references/product-definition-release.md 9.3 KB
- references/project-learning-loop.md 7.5 KB
- references/reference-image-learning.md 4.1 KB
- references/research-authority.md 8.5 KB
- references/zh-CN/autocad-mcp-workflow.md 18 KB
- references/zh-CN/complex-assembly-drafting.md 9.8 KB
- references/zh-CN/consumer-product-concept.md 12 KB
- references/zh-CN/drc-review.md 17 KB
- references/zh-CN/gbt-drafting.md 12 KB
- scripts/build_release_manifest.py 5.2 KB runs code
- scripts/catalog_reference_library.py 5.7 KB runs code
- scripts/check_dimension_authority.py 5.1 KB runs code
- scripts/check_document_identity.py 2.9 KB runs code
- scripts/check_interface_alignment.py 4.8 KB runs code
- scripts/check_model_manifest.py 4.5 KB runs code
- scripts/check_pre_cad_brief.py 4.2 KB runs code
- scripts/compare_render_viewset.py 7.5 KB runs code
- scripts/record_project_feedback.py 12 KB runs code
- scripts/validate_motion_states.py 4.5 KB runs code
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 12d ago First seen · 232 lines · 136 tokens per session scan A 8d0584e01697
industrial-product-design-gbt is a skill published in the GitHub repository beiming183-cloud/AutoCAD-MCP (19 stars, last pushed 1mo ago), licensed MIT. It adds 136 tokens to every session and 4,871 once invoked, about $0.0007 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
solidworks-vibecad
A planning guide for turning natural-language mechanical design requests into structured, parameterized SolidWorks plans that can be reviewed before execution.
kernelcad-kinematic
Use when verifying whether a moving assembly is buildable — sampled-pose collision sweeps across joint ranges, IK reachability for end-effector targets, mounting-hole fastener consistency, and static-load capacity on cantilever-shaped parts. Loads alongside kernelcad-authoring to gate design-time mechanism feasibility.
kernelcad-sheet-metal
Folded sheet-metal parts — sheetMetal(profile, opts), .bend(edge, angle, radius), .flattenPattern(). Use when building L-brackets, U-channels, service panels, or any part fabricated from a flat blank + folds.
21st-ui
Find, install, and generate UI with 21st.dev. Use when the user asks for a UI component (pricing table, hero, navbar, dashboard, form, etc.), wants design inspiration, needs a brand logo as an SVG component, or wants to generate new UI from a prompt.
autocad-automation
An automation skill for AutoCAD, a computer-aided design program commonly used for 2D technical drawings. It works with drawing formats such as DWG and DXF and can use Windows automation or other supported scripting routes.
solidworks-fillet-chamfer-cnc
A SolidWorks skill for creating CNC parts with multiple rounded edges and chamfers, which are angled or softened edges used in manufacturing. It includes parameter checks, edge selection, and evidence after rebuilding the model.