pfx
MCP Server für Forterro Proffix Px5 ERP - Verbinde Claude, ChatGPT und Gemini mit deinem Proffix Px5
claude mcp add --transport stdio pitwch-pfx-mcp-server node mcp-http-bridge.js https://mcp.pfx.ch/api/server \ --env PROFFIX_URL="https://dein-proffix.com" \ --env PROFFIX_PORT="11011" \ --env RESPONSE_FORMAT="json" \ --env PROFFIX_DATABASE="deine-datenbank" \ --env PROFFIX_PASSWORD="dein-passwort" \ --env PROFFIX_USERNAME="dein-benutzername" \ --env HTTP_AUTHORIZATION="Bearer YOUR_API_KEY_HERE"
How to use
The pfx MCP Server enables AI assistants to interact with Forterro Proffix Px5 data through the MCP (Model Context Protocol) API using JSON-RPC 2.0. It exposes Proffix tools such as searching endpoints, calling endpoints directly, and describing endpoints, allowing natural-language queries to be translated into precise Proffix API calls. Clients can request lists of available tools, initialize a session, and then issue tool calls with arguments. The server supports both JSON and TOON response formats, so results can be plain JSON for machine processing or enhanced with natural-language explanations for AI consumption. To get started, connect your MCP-enabled AI client (Claude, ChatGPT, Gemini, etc.) to the server’s API endpoint, authenticate with an API key, and begin issuing requests like: initialize, tools/list, and tools/call with the appropriate arguments.
How to install
Prerequisites:
- Node.js v18+ (recommended)
- Access to the pfx MCP server API key and Proffix Px5 credentials
- Internet access from the machine running the bridge
Installation steps:
- Clone the MCP server repository (or download the bridge file):
git clone https://github.com/pitwch/pfx-mcp-server.git
cd pfx-mcp-server
- Install dependencies (if any):
npm install
- Prepare configuration:
- Create a configuration file or use environment variables as shown in the example below.
- Ensure you have an API key and Proffix credentials ready.
- Run the bridge server:
# Example using the bridge file and environment variables
node mcp-http-bridge.js https://mcp.pfx.ch/api/server
Alternatively, set up as a background service with your OS service manager and export the needed environment variables (HTTP_AUTHORIZATION, PROFFIX_*, RESPONSE_FORMAT).
- Verify accessibility:
- Open the MCP server status endpoint or connect a client and issue initialize/tools/list to confirm connectivity.
Additional notes
Tips and common issues:
- Always use HTTPS for communication and avoid exposing API keys in client code.
- If you change Proffix credentials, update the corresponding environment variables and restart the bridge.
- Use RESPONSE_FORMAT=json for machine processing and RESPONSE_FORMAT=toon for AI-friendly responses.
- Ensure the HTTP_AUTHORIZATION header is kept secure; rotate API keys periodically.
- The server supports two MCP methods: initialize (handshake) and tools/list/call (operation discovery and execution). Use tools/list to discover available Proffix operations like proffix_search_endpoints, proffix_call_endpoint, and proffix_describe_endpoint.
- When debugging, enable verbose logs on the bridge and verify network paths between the bridge host and your Proffix Px5 instance.
Related MCP Servers
zen
Selfhosted notes app. Single golang binary, notes stored as markdown within SQLite, full-text search, very low resource usage
MCP -Deepseek_R1
A Model Context Protocol (MCP) server implementation connecting Claude Desktop with DeepSeek's language models (R1/V3)
mcp-fhir
A Model Context Protocol implementation for FHIR
mcp
Inkdrop Model Context Protocol Server
mcp-appium-gestures
This is a Model Context Protocol (MCP) server providing resources and tools for Appium mobile gestures using Actions API..
dubco -npm
The (Unofficial) dubco-mcp-server enables AI assistants to manage Dub.co short links via the Model Context Protocol. It provides three MCP tools: create_link for generating new short URLs, update_link for modifying existing links, and delete_link for removing short links.
