Loading languages...
PR

proxymock

@Speedscale

An MCP server is built into proxymock to let Agent-based LLM tools like Cursor, Claude Desktop, VSCode or GitHub Copilot interact.

API mocking
testing
developer tool
LLM integration

proxymock is a tool that records API/Database calls from an app running on your local desktop. It automatically converts inbound requests into tests and outbound requests into service mocks, with portable recordings.

An MCP server is integrated into proxymock, allowing agent-based LLM tools such as Cursor, Claude Desktop, VSCode, and GitHub Copilot to interact with it. This enables users to control proxymock actions (like starting/stopping mock servers, recording API calls, or running regression tests) directly through their normal chat interfaces.

Installation and Usage:

  1. Install proxymock (instructions available on proxymock.io or docs.speedscale.com).
  2. The proxymock init command typically auto-detects common MCP clients and prompts for installation.
  3. If manual configuration is needed, modify your client's mcp.json with the provided configuration:
    {
        "mcpServers": {
            "proxymock-simulator": {
                "command": "/Users/some_user/.speedscale/proxymock",
                "args": [
                    "mcp"
                ]
            }
        }
    }
    
  4. Alternatively, run proxymock mcp --install to re-detect and install.

Example Prompts for LLM Agents:

  • run a mock server
  • stop the mock server
  • start recording api calls
  • run a regression test

# mcpServer Config

{
  "mcpServers": {
    "proxymock-simulator": {
      "command": "/Users/some_user/.speedscale/proxymock",
      "args": [
        "mcp"
      ]
    }
  }
}

# stdio

proxymock mcp
Transport:
stdio
Language:
Updated: 7/31/2025