You can build your “Local” MCP Server on the Peliqan Data Cloud, in 3 steps:
- In Peliqan, build a custom API endpoint
- Add an app of type “API handler” and add the below example script.
- Create an API endpoint with method POST and path
/mcp
, linked to your API handler script. Click here for more info on how to publish an API endpoint in Peliqan. - Install the Peliqan MCP Server on your computer
- Configure the Peliqan MCP Server in your MCP Client, for example in Claude Desktop
The deployment scenario is also called “Bridged MCP”, because the locally installed MCP Server acts as a bridge or proxy to methods that you define on the Peliqan Data Cloud.
You can write, update and maintain your MCP Server functions (tools) in the cloud on Peliqan, without the need for end-users to update their local installation ! The proxy handles dynamic discovery of new functions (tools):
If your MCP Client supports “Remote MCP”, see the section Remote MCP Server on Peliqan instead.
1. Set up a custom API endpoint in Peliqan
Add a custom API endpoint in Peliqan, and use the below script as the API handler. In the example script, replace the functions decorated with @mcp.tools()
with your own MCP functions (tools).
2. Install the Peliqan MCP Server on your computer
Install the Peliqan MCP Server on your computer:
pip install mcp-server-peliqan
See Peliqan MCP Server on Github.
3. Configure your MCP Client - example for Claude Desktop
In order to use your MCP Server in Claude Desktop, create a file claude_desktop_config.json
.
Location on Windows: $env:AppData\Claude\claude_desktop_config.json
Location on Mac: ~/Library/Application\ Support/Claude/claude_desktop_config.json
Test in Claude Desktop
Restart Claude Desktop. Next click on the “Search & tools” icon in the bottom left corner to see the installed MCP servers, double check if Peliqan is listed and that is shows the tools that you implemented in your MCP Server.