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 agentmods add instructions/danielproxd2/solidworks-mcp/claude-mdgit clone --depth 1 https://github.com/danielproxd2/solidworks-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/instructions/danielproxd2/solidworks-mcp/claude-md)<a href="https://agentmods.dev/instructions/danielproxd2/solidworks-mcp/claude-md"><img src="https://agentmods.dev/badge/instructions/danielproxd2/solidworks-mcp/claude-md.svg" alt="Measured on agentmods" 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 | $0.01111 | $0.01111 |
| Opus 5 | $0.00556 | $0.00556 |
| Sonnet 5 | $0.00222 | $0.00222 |
| Haiku 4.5 | $0.00111 | $0.00111 |
Grade A, and why
solidworks-mcp CLAUDE.md 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 5d 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 — 79 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Repository guidance
Operating rules for automated contributors and maintainers working in this repository.
Project scope
MCP_CAD exposes a constrained MCP interface for controlled SOLIDWORKS automation. The core server is client- and model-provider agnostic.
Safety constraints
- Design files remain local. Do not upload or commit
.sldprt,.sldasm,.slddrw, STEP files, captures derived from private designs, or other user geometry. - Use the constrained tool surface. Automation must call predefined, deterministic operations rather than arbitrary SOLIDWORKS API methods.
- Require explicit approval for model changes. Do not mutate geometry without the user's approval.
- Protect credentials. Keep API keys in user-controlled configuration; never log, transmit, or commit them.
- Do not collect design telemetry. Avoid analytics containing model contents, file names, geometry patterns, or derived design data.
Architecture
src/mcp_cad/server.pydefines FastMCP tools and binds them to the module-level client.src/mcp_cad/solidworks.pydefines theSolidWorksClientProtocol, the in-memory mock, and the Windows pywin32 COM implementation.MCP_CAD_USE_MOCK=1selects the mock for development on systems without SOLIDWORKS.
New geometry behavior should be added end to end: Protocol, mock implementation, COM implementation, then MCP tool. Keep pywin32-specific code out of server.py. Advisory and composite operations may live only in server.py when they do not require a new Protocol method.
The live tool catalog and tool docstrings are authoritative. Tool descriptions consume MCP context on every tools/list, so keep them concise and keep orchestration guidance in repository documentation.
Build discipline
- Prefer the currently open document. Read its state first, and create a new document only when requested or necessary.
- Use millimeters and degrees above the COM boundary. Convert millimeters to meters only inside
WinComSolidWorksClient. - SOLIDWORKS COM mutation is serial. Never run geometry-mutating calls concurrently.
- Issue one mutator at a time unless using an explicit batch or plan runner.
- Treat each mutation as at most once. After a lost or ambiguous response, inspect live state before deciding whether anything should be retried.
- Use
receipt.feature_countas the normal post-mutation signal. Perform separate reads when receipts are missing or unexpected. - Save early and again after verification.
- Prefer reversible suppression or modification over deletion. Do not add destructive behavior casually.
- Reuse an indexed part or edit a driving dimension when either is appropriate; do not rebuild existing geometry unnecessarily.
- Never invent missing dimensions, tolerance classes, materials, or standards. Ask a focused question or report the unresolved requirement.
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.
- 5d ago First seen · 79 lines · 1,111 tokens per session scan A 222d886bfda4
solidworks-mcp CLAUDE.md is an instructions file published in the GitHub repository danielproxd2/solidworks-mcp (1 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 1,111 tokens to every session, about $0.0056 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 2026-08-31.
Other instructions, from other repositories
hass-mcp-server AGENTS.md
AGENTS.md instructions for ganhammar/hass-mcp-server, covering working on this integration, calling into home assistant, before replacing a private call with a service, anything left in category 2 or 3 and conventions.
gfty AGENTS.md
Instructions for rkana-org/gfty, covering gfty development guide, scope and purpose, repository map, core behavior and invariants and configuration schemas and products.
horizun-revit-mcp AGENTS.md
AGENTS.md instructions for HorizunGroup/horizun-revit-mcp, covering horizun revit mcp — guide for agents / guía para agentes, english, install, configure the mcp client and claude code — user scope makes it available across projects.
cad-cae-copilot copilot-instructions.md
Copilot instructions for armpro24-blip/cad-cae-copilot, covering github copilot — aieng workspace and essentials.
gpu-mcp-server AGENTS.md
Instructions for pmady/gpu-mcp-server, covering agents.md, project overview, build, code layout and key patterns.
pixoo-mcp-server AGENTS.md
Instructions for cyanheads/pixoo-mcp-server, covering developer protocol, architecture overview, what's next?, core rules and patterns.