mcp-agent

Open-source framework for building agents on the Model Context Protocol

Editorially reviewedChecked Sep 2026How we review

mcp#model-context-protocol#agent-framework#open-source#workflow-patterns

Free plan API Self-hosted

Toolglade’s take

Shines if you are all-in on Model Context Protocol and want composable agent workflow patterns plus managed server lifecycles and optional Temporal durability. The caveat is it is Python-only and its usefulness rises and falls with MCP ecosystem maturity.

Should you use mcp-agent?

Best for: Developers building on MCP · Teams needing composable agent patterns · MCP server integrators
Not ideal for: No-code users · Teams not using MCP · Non-Python stacks
Biggest strength: Purpose-built for MCP
Biggest weakness: Python-only
Pricing: Free plan, paid from $0
Top alternative: FastMCP · see all

Facts & pricing last checked September 2026. How we review

About mcp-agent

mcp-agent is an open-source Python framework from LastMile AI for building composable, model-agnostic agents on the Model Context Protocol using proven workflow patterns.

mcp-agent, from LastMile AI, is a lightweight framework that makes it practical to build agents on top of the Model Context Protocol. It introduces an AugmentedLLM, an LLM enhanced with tools from a collection of MCP servers, and implements every workflow pattern from Anthropic's Building Effective Agents plus the OpenAI Swarm pattern. Because each pattern is itself an AugmentedLLM, patterns compose and chain cleanly, keeping agent logic modular and model-agnostic. The framework provides full MCP client support, including tools, resources, prompts and notifications, plus advanced features like OAuth, sampling, elicitation and roots, and it supports persistent state through durable execution. MCP is low-level, and mcp-agent handles the mechanics of server connections, LLM interaction and human-in-the-loop signals so developers can focus on business logic. Documentation, an SDK and CLI are available at docs.mcp-agent.com, making it a strong open-source choice for teams standardizing on MCP.

Weighing your options?See how mcp-agent compares to the alternatives.

TL;DR

mcp-agent is LastMile AI's open-source Python framework for building composable, MCP-native agents using proven workflow patterns.

Company overview

mcp-agent is an open-source project from LastMile AI focused on building agents on the Model Context Protocol. It targets developers who want to leverage MCP servers and standard agent patterns without wiring up low-level plumbing.

The project provides an SDK, CLI and documentation, positioning itself as a practical, MCP-first framework in a growing ecosystem of MCP tooling.

Product features

The framework centers on the AugmentedLLM, an LLM enhanced with MCP server tools, and implements composable workflow patterns from Anthropic's Building Effective Agents and OpenAI Swarm. It offers full MCP client support including tools, resources, prompts, OAuth, sampling and elicitation.

Durable execution enables persistent state and human-in-the-loop signals, so developers can build reliable, stateful agents while the framework handles MCP mechanics.