AICraft MCP Server
This server has a remotely hosted endpoint. Open its connector listing for connection details, authentication requirements, and setup instructions:
AICraft MCP Server
A Model Context Protocol server that gives AI agents access to injection molding process optimization: first-shot process recommendation and defect-driven tuning.
This is a thin MCP client. All calculation, authentication, and billing happen on the hosted HTTPS API. No process data or algorithm code runs locally.
What it does
Injection molding engineers need two things: a sensible starting set of machine parameters before a trial run, and corrections when the trial shows defects. This server exposes both as MCP tools, backed by a self-developed rule engine with explainable reasoning.
Tool | What it does | Cost |
| Free catalog of fields, units, defect codes, and error codes | Free |
| Recommend initial process parameters from material, part, and machine data | 1 API call |
| Adjust current parameters for an observed defect | 1 API call |
The paid tools require a subscription API key. Get one at https://aicraft4injection.com ($19/month, 1000 calls). The schema tool works without a key so agents can discover the contract before paying.
Related MCP server: mcp-design-platform
Requirements
Python 3.10+
An API key from https://aicraft4injection.com (for paid tools)
Installation
pip install -r requirements.txtConfiguration
Set two environment variables:
export AICRAFT_API_KEY="sk_your_key_here" # required for paid tools
export AICRAFT_BASE_URL="https://aicraft4injection.com" # optional, this is the defaultUsage with MCP clients
Add this server to your MCP client configuration (Claude Desktop, Cursor, Cline, etc.):
{
"mcpServers": {
"aicraft": {
"command": "python",
"args": ["/absolute/path/to/agent-mcp-server/server.py"],
"env": {
"AICRAFT_API_KEY": "sk_your_key_here",
"AICRAFT_BASE_URL": "https://aicraft4injection.com"
}
}
}
}Restart the client. Three tools should appear: schema, first_shot, optimize.
Tool reference
schema
No arguments. Returns field groups, default units (mm, mm/s, MPa, rpm, °C, s), valid defect codes, and error codes.
first_shot
Requires three fields (the engine cannot compute without them):
Field | Meaning |
| Material short name, e.g. |
| Average wall thickness in mm (0 < value < 50) |
| Screw diameter in mm (> 0) |
Recommended but optional: melt density, back pressure, drying temperature, gate thickness, clamp force, injection stroke, barrel temperature zones.
Example argument:
{
"polymer": {"polymer_abbreviation": "PC", "melt_density": 1.05},
"mold": {"product_ave_thickness": 2.0, "gate_thickness": 1.2},
"machine": {"machine_clamp_force_ton": 180,
"injector": {"injector_screw_diameter_mm": 36}}
}optimize
Requires a defect and the current process parameters:
{
"defect_info": [{"desc": "FLASH", "level": "moderate"}],
"current_params": {"IP0": 90, "IV0": 40, "PP0": 50, "CT": 12}
}desc accepts stable codes such as SHORTSHOT, FLASH, SHRINKAGE, BURN, WELDLINE, WARPING (25 codes in total — call schema for the full list).
level accepts an English enum or a number from 0 to 1 (no Chinese values):
Value | Meaning |
| No defect — the row is dropped |
| Minor defect |
| Medium defect |
| Serious defect |
| Very serious defect |
The gateway translates these to the engine's internal severity values. Invalid levels are rejected with 400 INVALID_FIELD and do not consume quota.
Design notes
Stdio transport, protocol version
2024-11-05for broad client compatibility.No local computation. The server forwards JSON-RPC tool calls to the hosted API; the engine stays on the operator's own infrastructure.
Stable error codes. Every API response carries an
api_versionand a machine-readablecode(e.g.MISSING_FIELD,UNKNOWN_DEFECT,QUOTA_EXCEEDED,ENGINE_REJECTED). Agents branch oncode, not on prose.No charge on invalid input. Requests rejected by validation do not consume quota. Engine failures are refunded automatically.
A Streamable HTTP endpoint is also available for clients that prefer remote MCP: https://aicraft4injection.com/mcp (discovery document at /.well-known/mcp.json).
Links
Service: https://aicraft4injection.com
Free field/defect catalog: https://aicraft4injection.com/v1/schema
Short OpenAPI spec: https://aicraft4injection.com/openapi-tools.json
LLM discovery file: https://aicraft4injection.com/llms.txt
License
MIT
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
This server cannot be deployed
Related MCP Connectors
Manufacturing capability graph for AI agents: materials, processes, tolerances, and instant quotes.
Hosted MCP for e-commerce: live product catalog, stock, and pricing for AI agents.
An MCP gateway for teams deploying AI tools at scale to employee-facing and headless agents.
Compliance frameworks (SOC 2, ISO 27001, CMMC, NIST, more) delivered to AI agents as MCP tools.
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceMCP server for integrating manufacturing systems (MES/ERP/quality/maintenance) with LLM agents, enabling event ingestion, incident triage, approval workflows, and RAG-based knowledge retrieval.MIT
- AlicenseNot gradedqualityDmaintenanceEnables AI agents to remotely operate visual design tools via MCP protocol, with composable architecture for image processing, file operations, and workflow automation.MIT
- FlicenseNot gradedqualityDmaintenanceEnables AI assistants to inspect and automate Autodesk Fusion 360 through MCP, providing tools for CAD modeling, CAM manufacturing, and document management.-
- FlicenseNot gradedqualityDmaintenanceEnables agents to inspect machine status and sensor anomalies, search maintenance guidance, and create maintenance tickets through controlled MCP tools.-