Dev.to AI 🤖 Ai 👁 0 📖 2 min read

A free MCP server for engineering materials data, with a source for every number

Ask an AI assistant for the yield strength of 316L and you usually get a number with no source. Material properties are scattered across data sheets, standards and papers, and most numbers lose their test conditions on t

Ask an AI assistant for the yield strength of 316L and you usually get a number with no source. Material properties are scattered across data sheets, standards and papers, and most numbers lose their test conditions on the way.

I run Space Mushroom, a free open materials database where every value is stored with its unit, test conditions, method and a pointer to the source. I have put it behind a read-only MCP server, so an assistant can look a value up and show where it came from.

Connect it

No API key and no sign-in. Endpoint (Streamable HTTP): https://spacemushroom.com/mcp

Claude Code:

claude mcp add --transport http space-mushroom https://spacemushroom.com/mcp

Cursor, VS Code and other clients that accept a remote server:

{
  "mcpServers": {
    "space-mushroom": { "url": "https://spacemushroom.com/mcp" }
  }
}

It is also listed in the official MCP registry as com.spacemushroom/materials.

What it does

Eleven tools, all read-only. The ones I reach for most:

  • find_materials: requirements in, candidates out. Each material gets pass, fail or unknown, with the evidence behind every constraint. Unknown matters: "no data" is not the same as "fails".
  • compare: line materials up on one property, each value with its conditions and source.
  • prove_claim: everything behind one value, as printed in the source, converted, the conditions, and where in the source it sits.
  • resolve_designation: UNS, EN, DIN, JIS, AISI and ASTM codes, and trade names, to the grade.
  • what_is_missing: which properties have data for a material and which have none yet.

An example

Compare the density of aluminium 6061 and Ti-6Al-4V and tell me where each number comes from.

compare returns 2.7 g/cm3 for 6061 (an Aluminum Association typical value for the T6 and T651 tempers, from an ASM data sheet) and 4.43 g/cm3 for Ti-6Al-4V (from the physical-constants section of a technical reference). Both are marked ai_checked: an AI assistant re-read them against the source and a person has not. That label travels with the value.

Choices I made on purpose

  • A missing value means "not covered yet", never "safe".
  • Different test conditions are listed separately, not averaged.
  • The data is CC BY 4.0 and every answer includes the source to cite.
  • The server only reads. It uses a public key, and the database's row-level security is the access control.

Links

It is young, so some materials and properties are missing. If you try it and cannot find one you need, tell me and I will add it.

📰 Read the original article on Dev.to AI

Originally published by Dev.to AI. Aggregated on AIWithGhost for educational purposes — full credit and traffic to the original publisher.