Pearl MCP Server

Name: Pearl MCP Server
Author: Pearl-com

@Pearl-com

About Pearl MCP Server

A Model Context Protocol (MCP) server implementation that exposes Pearl's AI and Expert services through a standardized interface

#mcp #human-in-the-loop #human-verification #expert-advisor #agentic-ai #mcp-server

Basic information

Config

Add this server to your MCP-compatible client using the configuration below.

{
  "mcpServers": {
    "pearl_mcp_server": {
      "command": "python",
      "args": [
        "-m",
        "venv",
        ".venv"
      ]
    }
  }
}

Tools

assisted human expert support

Same as ask_pearl_ai

`session_id`

Overview

What is Pearl MCP Server?

A Model Context Protocol (MCP) server implementation that exposes Pearl’s AI and Expert services through a standardized interface. It allows MCP clients such as Claude Desktop, Cursor, and other MCP‑compatible applications to interact with Pearl’s advanced AI assistants and human experts.

How to use Pearl MCP Server?

Install Python 3.12 or higher, obtain a Pearl API key from the Pearl contact page, clone the repository, create a virtual environment, install dependencies with pip install -e ., and set the PEARL_API_KEY environment variable in a .env file. Run the server locally with pearl-mcp-server --api-key your-api-key (stdio by default) or with --transport sse. Alternatively, connect directly to the hosted server at https://mcp.pearl.com/mcp without local installation. Configure your MCP client (e.g., Claude Desktop, Cursor, Windsurf) with the appropriate JSON settings for stdio or use mcp-remote as a bridge for remote connections.

Key features of Pearl MCP Server

Support for both stdio and SSE transports
Integration with Pearl API for AI and expert assistance
Session management for continuous conversations
Two interaction modes: AI‑Expert and direct Expert
Conversation history tracking and stateful session management
Automatic routing of queries to the most relevant expert category

Use cases of Pearl MCP Server

Getting AI‑assisted human expert support for complex or sensitive topics
Obtaining direct human expert advice in medicine, law, finance, technology, education, and lifestyle
Checking the status and history of an ongoing expert conversation
Building custom Python clients that leverage Pearl’s AI and expert services via MCP

FAQ from Pearl MCP Server

What transports does Pearl MCP Server support?

It supports both stdio (default) and SSE transports. SSE can be selected with the --transport sse flag and a custom port.

How do I obtain a Pearl API key?

Visit the Pearl Contact Page and request an API key for MCP server integration. Keep your API key secure and never commit it to version control.

What are the prerequisites for running the server locally?

Python 3.12 or higher, a Pearl API key, and either the pip or uv package manager.

Can I use the server without installing it locally?

Yes. Pearl provides a hosted MCP server at https://mcp.pearl.com/mcp that can be used directly with any MCP client, optionally via the mcp-remote package.

What tools are available on this server?

The server provides four tools: ask_pearl_expert (AI‑assisted human expert), ask_expert (direct human expert), get_conversation_status, and get_conversation_history.