llama-index-tools-mcp

What it is and what it does

This package bridges LlamaIndex agents and the Model Context Protocol (MCP) ecosystem. It provides a client to connect to MCP servers, extract their exposed tools, and make those tools available to LlamaIndex agents for use in agentic workflows. It also works in reverse: it can wrap a LlamaIndex Workflow as an MCP server so other MCP clients can call it.

The package handles the mechanics of MCP communication—tool listing, resource access, prompt retrieval, and OAuth authentication—so you can focus on integrating external capabilities into your agent. It supports both synchronous and asynchronous operations, multiple transport types (HTTP, Server-Sent Events, stdio), and optional filtering of tools by name.

Use it for:

• Equip a LlamaIndex agent with tools from an external MCP server (e.g., weather, database, file system).
• Expose a LlamaIndex workflow as an MCP service for other applications or agents to call.
• Build an agent that queries multiple MCP servers and combines their tools in a single reasoning loop.
• Implement OAuth-protected access to MCP servers requiring authentication.
• Convert a complex LlamaIndex workflow into a reusable tool for other systems via MCP.

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

Connects LlamaIndex agents to MCP (Model Context Protocol) servers to expose and call remote tools, and converts LlamaIndex workflows into MCP servers.

Yes. If you are building LlamaIndex agents and need to integrate tools from MCP servers, this package eliminates the integration work. It has low install friction, active maintenance, no known vulnerabilities, and a permissive license. Install it when you have an MCP server available and want agents to use its tools.

Install

llama-index-tools-mcp on PyPI

pip install llama-index-tools-mcp

uv add llama-index-tools-mcp

poetry add llama-index-tools-mcp

Installing llama-index-tools-mcp

Before you install

Low install friction with three straightforward runtime dependencies. Active maintenance with a recent release.

License in practice

MIT license permits unrestricted use, modification, and distribution with minimal obligations.

Quickstart

from llama_index.tools.mcp import BasicMCPClient, McpToolSpec

mcp_client = BasicMCPClient("http://127.0.0.1:8000/sse")
mcp_tool_spec = McpToolSpec(client=mcp_client)
tools = mcp_tool_spec.to_tool_list()

Requires an MCP server running and accessible at the specified URL or command.

Verify before relying

• Whether BasicMCPClient supports all MCP transport types mentioned (HTTP, SSE, stdio) equally well in practice
• Performance characteristics when calling tools through multiple MCP servers simultaneously
• Compatibility guarantees with specific MCP server implementations or versions

Package facts

Evidence: llama_index_tools_mcp-0.5.0-py3-none-any.whl

Keywords: mcp, tools

Tags

More Artificial Intelligence packages

Further reading