About CutOptim Engine
CutOptim Engine is an MCP server for AI assistants that runs cutting optimization through the CutOptim Engine API, covering panels, bars, timber, and irregular shapes. It supports both a hosted endpoint and a local stdio package.
Key features
- 012D rectangular panel layouts
- 02Bar, tube, and profile optimization
- 03Timber cross-section optimization
- 04Free-form polygon nesting
- 05Request validation and job lookup
Use cases
- 01Plan panel layouts and cutting plans
- 02Optimize stock lengths for bars or tubes
- 03Arrange timber parts by cross-section
- 04Validate nesting requests and inspect async jobs
MCP support
The MCP server exposes CutOptim Engine API optimization, validation, job, and usage functions as tools. MCP clients can connect to the hosted endpoint or use the local stdio package.
Configuration
Remote servers need no local install. Pick your client, then copy the config or install in one click.
https://api.cutoptim.com/mcpclaude mcp add --transport http cutoptim-engine https://api.cutoptim.com/mcp{"mcpServers":{"cutoptim-engine":{"type":"http","url":"https://api.cutoptim.com/mcp"}}}Open in Cursor{"servers":{"cutoptim-engine":{"type":"http","url":"https://api.cutoptim.com/mcp"}}}Open in VS Code{"mcpServers":{"cutoptim-engine":{"type":"http","url":"https://api.cutoptim.com/mcp"}}}{"mcpServers":{"cutoptim-engine":{"serverUrl":"https://api.cutoptim.com/mcp"}}}Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.
https://api.cutoptim.com/mcpAuth requiredTools
Details
From the project READMECutOptim MCP server
Cutting optimization as tools for AI assistants: 2D panels, 1D bars, timber and true-shape nesting,
through the CutOptim Engine API and the Model Context Protocol.
MCP guide · API reference · Get a free test key
Two ways to connect
Hosted (no install) — point your MCP client at https://api.cutoptim.com/mcp with the header
Authorization: Bearer <your API key>. Configurations for Claude Code, Cursor, VS Code and Claude Desktop are in the
MCP guide.
Local (this package, stdio) — runs on your machine and calls the same API with your key:
{
"mcpServers": {
"cutoptim": {
"command": "npx",
"args": ["-y", "@cutoptim/mcp"],
"env": { "CUTOPTIM_API_KEY": "co_test_YOUR_KEY" }
}
}
}
Claude Code: claude mcp add cutoptim --env CUTOPTIM_API_KEY=co_test_YOUR_KEY -- npx -y @cutoptim/mcp
| Variable | Meaning |
|---|---|
CUTOPTIM_API_KEY | your key: co_live_…, co_test_… (never billed) or the public demo key. Without one, only the validate_* tools work. |
CUTOPTIM_BASE_URL | default https://api.cutoptim.com |
Node 18 or newer.
Tools
| Tool | What it does | Cost |
|---|---|---|
optimize_2d | rectangular parts from sheets: layout, cut plan, yield | one API call |
optimize_1d | lengths from bars, tubes, profiles | one API call |
optimize_wood | timber parts by cross-section | one API call |
optimize_nest | free-form polygon parts on sheets | one API call |
validate_2d · validate_1d · validate_wood · validate_nest | check a request without solving it | free |
get_job | read an async engine: "max" job | free |
get_usage | calls used and left this month | free |
The arguments of each tool are exactly the request body of its endpoint — the schemas are generated from the
OpenAPI document. Every result carries a short summary for the assistant
and the API's JSON answer as structuredContent.
Quota and billing
One tool call is one Engine API call: the same key, quota, rate limits and billing as calling the REST API yourself.
The client sends an Idempotency-Key with every optimize call, so its own retries are never charged twice.
License and support
MIT. This repository is published from the CutOptim monorepo; report a bug in Issues or write to [email protected].
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.
