Command Palette

Search for a command to run...

UnylyUnyly
Весь каталог

Outgassing

БесплатноНе проверен

MCP server for querying outgassing properties of materials through approved aerospace sources.

GitHubEmbed

Описание

MCP server for querying outgassing properties of materials through approved aerospace sources.

README

License: Apache 2.0 Python 3.14+ MCP Docker

A Model Context Protocol (MCP) server that provides access to NASA's outgassing database for aerospace material selection. This server enables querying of material outgassing properties (TML, CVCM, WVR) by compatible large language models.

Demo

Araldite 2014 Demo

Overview

This MCP server is built with FastMCP and containerized with Docker. It provides four tools for querying 13,582 aerospace materials from NASA's outgassing database, with intelligent fuzzy search and automatic compliance checking against NASA-STD-6016 standards.

Features

  • Fuzzy Material Search: Find materials by name using intelligent matching (rapidfuzz)
  • Application-Based Search: Filter materials by usage type (ADHESIVE, POTTING, TAPE, etc.)
  • Compliance Checking: Automatic TML/CVCM validation with WVR adjustment
  • VS Code Integration: Seamless access through GitHub Copilot Chat
  • Corporate Network Support: Built-in Zscaler certificate handling

Tools

1. query_materials

Search for materials by name with fuzzy matching and compliance validation.

Parameters:

  • material (string): Material name to search for
  • max_tml (float, optional): Maximum acceptable TML % accounting for WVR if present (default: 1.0)
  • max_cvcm (float, optional): Maximum acceptable CVCM % (default: 0.1)
  • limit (int, optional): Maximum number of results to return (default: 10)

Returns: JSON string with materials matched in the database, sorted by match score (best first, 100 being exact match, less than 82 being a low quality match)

Example usage in Copilot Chat:

Check the outgassing-mcp-server for RTV adhesives with TML under 0.5%.

2. get_material

Get detailed material information by unique ID from the outgassing database.

Integration pattern: Call query_materials() or query_application() to find material IDs before invoking this function.

Parameters:

  • ID (string): Unique material ID in the outgassing database

Returns: JSON string with complete material details or error message if ID not found

Example usage in Copilot Chat:

Get details for material ID GSC32950 from the outgassing-mcp-server.

3. get_applications

Retrieve all unique application/usage types from the database.

Parameters: None

Returns: JSON string with list of all available application categories (961 unique types)

Example usage in Copilot Chat:

Check the outgassing-mcp-server to see what material application types are listed?

4. query_application

Query materials by application/usage meeting specified TML and CVCM limits.

Integration pattern: Call get_applications() to retrieve available application/usage types before invoking this function.

Parameters:

  • application (string): Application/usage type to search for (e.g., ADHESIVE, POTTING, TAPE)
  • max_tml (float, optional): Maximum acceptable TML % (default: 1.0)
  • max_cvcm (float, optional): Maximum acceptable CVCM % (default: 0.1)

Returns: JSON string with materials meeting application and outgassing criteria sorted by adjusted TML with lowest values first

Example usage in Copilot Chat:

Check the outgassing-mcp-server to find compliant adhesives for spacecraft use.

Understanding Outgassing Data

  • TML (Total Mass Loss): Percentage of material mass lost in vacuum
  • CVCM (Collected Volatile Condensable Material): Percentage that condenses on a collector
  • WVR (Water Vapor Recovery): Mass of water vapour regained by specimen following outgassing
  • NASA-STD-6016: [MPR 95] Requires TML ≤ 1.0% and CVCM ≤ 0.1% (after WVR adjustment)
  • Adjusted TML: TML - WVR Used when absorbed water lost does not affect functionality
  • Special Cases: Contamination sensitive surfaces such as optical lenses may require a lower CVCM than the above

Prerequisites

  • Docker: Ensure Docker is installed and running on your system
  • VS Code: Latest version with GitHub Copilot extension enabled
  • Network Access: Internet connectivity for downloading dependencies (corporate networks may require additional certificate setup)

Installation & Setup

1. Clone the repository locally

Navigate to the location you wish to store the repository and clone it:

git clone https://github.com/aerospace-mcp-tools/outgassing-mcp-server.git

1. Certificate Setup (Corporate Networks Only)

If you're behind a corporate firewall using Zscaler or similar proxy:

  1. Obtain your organization's root certificate
  2. Copy the certificate file to the repository root
  3. Rename it to zscaler-root-ca.crt

See SECURITY.md for using Docker with Zscaler

2. Build the Docker Image

Navigate to the repository directory and build the Docker image:

docker build -t outgassing-mcp-server .

This command:

  • Downloads Python 3.14 base image
  • Installs required dependencies using UV package manager
  • Configures Docker for Zscaler
  • Sets up the FastMCP server environment

3. Verify Installation

Test that the server builds and runs correctly:

docker run -it --name test outgassing-mcp-server

You should see the FastMCP startup banner:

FastMCP Banner

Stop the test container and clean up:

docker stop test
docker rm test

VS Code Integration

This MCP server works with any MCP-compatible client that supports stdio transport. Each client will have its own configuration file location and JSON structure, but all use the same Docker container and command. The example given below shows integration with Github Copilot on VS Code.

Method 1: Using VS Code Command Palette (Recommended)

  1. Open VS Code and press Ctrl+Shift+P (Windows/Linux) or Cmd+Shift+P (macOS)
  2. Type and select "MCP: Add Server"
  3. Choose "Command (stdio)" as the transport type
  4. Enter the Docker command:
    docker run --rm -i --network host outgassing-mcp-server
    
  5. Name the server: outgassing-mcp-server
  6. The server will be automatically added to your MCP configuration

Method 2: Manual Configuration

Alternatively, you can manually edit your MCP configuration file:

Location: %APPDATA%\Code\User\mcp.json (Windows) or ~/.config/Code/User/mcp.json (Linux/macOS)

Add the following configuration:

{
  "servers": {
    "outgassing-mcp-server": {
      "type": "stdio",
      "command": "docker",
      "args": [
        "run",
        "--rm",
        "-i",
        "--network",
        "host",
        "outgassing-mcp-server"
      ]
    }
  }
}

Verification

  1. Restart VS Code after configuration
  2. Open GitHub Copilot Chat
  3. Test the mcp server with an example from above!

Data Source

Configuration Details

Docker Run Parameters Explained

  • --rm: Automatically remove container when it exits (prevents container buildup)
  • -i: Keep STDIN open for interactive communication with MCP protocol
  • --network host: Use host networking for seamless VS Code stdio communication
  • outgassing-mcp-server: The Docker image name built in earlier steps

Development & Contributing

Interested in contributing to this project? Please see CONTRIBUTING.md for:

  • Development setup instructions
  • Testing procedures
  • Code guidelines
  • How to submit changes

Project Structure

outgassing-mcp-server/
├── main.py                       # FastMCP server 
├── Dockerfile                    # Multi-stage build with cert support
├── pyproject.toml                # Dependencies (fastmcp, pandas, rapidfuzz)
├── data/
|   └── Outgassing_Db_rows.csv    # Local data cache (12,859 materials)
├── README.md                     # This file
├── CONTRIBUTING.md               # Contribution guidelines
├── SECURITY.md                   # Security policy and considerations
└── .github/
    └── copilot-instructions.md  # Development guidelines for LLMs

Troubleshooting

Container Issues

Stale data issues:

  • Outgassing data is only downloaded when container is built
  • Re-build container to get latest data

Container fails to start:

  • Verify Docker is running: docker --version
  • Check image exists: docker images | grep outgassing-mcp-server
  • Review build logs for errors during image creation
  • Test manually: docker run -it --name test outgassing-mcp-server

VS Code Integration Issues

VS Code doesn't detect the server:

  • Confirm mcp.json configuration syntax is valid JSON
  • Verify file location: %APPDATA%\Code\User\mcp.json (Windows)
  • Restart VS Code after configuration changes
  • Check VS Code Developer Console (Help > Toggle Developer Tools) for MCP-related errors
  • Ensure Docker Desktop is running before starting VS Code

Server appears but tools don't work:

  • Rebuild the Docker image to ensure latest code
  • Check that FastMCP banner shows on startup
  • Test tools manually: docker run -it --name test outgassing-mcp-server

Technical Details

Dependencies

  • Python: 3.14 (slim-trixie base image)
  • FastMCP: ≥2.13.0.2 (MCP protocol implementation)
  • pandas: ≥2.3.3 (data processing)
  • rapidfuzz: ≥3.14.3 (fuzzy string matching)

Security Considerations

For detailed security information, including vulnerability reporting, and privacy considerations, please see SECURITY.md.

Support & Resources

Project Documentation

Acknowledgments

  • Data Source: NASA Outgassing Database
  • Framework: FastMCP by the FastMCP team
  • Protocol: Model Context Protocol (MCP) specification

from github.com/aerospace-mcp-tools/outgassing-mcp-server

Установка Outgassing

У этого сервера нет опубликованного пакета — он собирается из исходников. Открой репозиторий и следуй инструкции в README.

▸ github.com/aerospace-mcp-tools/outgassing-mcp-server

FAQ

Outgassing MCP бесплатный?

Да, Outgassing MCP бесплатный — установка в пару кликов через Unyly без оплаты.

Нужен ли API-ключ для Outgassing?

Нет, Outgassing работает без API-ключей и переменных окружения.

Outgassing — hosted или self-hosted?

Self-hosted: сервер запускается локально на твоей машине командой из раздела установки.

Как установить Outgassing в Claude Desktop, Claude Code или Cursor?

Открой Outgassing на unyly.org, выбери вкладку своего клиента (Claude Desktop, Claude Code, Cursor) и нажми Install — конфиг сгенерируется автоматически, без правки JSON.

Похожие MCP

Compare Outgassing with

Не уверен что выбрать?

Найди свой стек за 60 секунд

Автор?

Embed-бейдж для README

Похожее

Все в категории development