mcpField
M
MCP Agent Strategy

MCP Agent Strategy

github.com/hjlarryFrameworks & SDKs17Updated 82w ago
Repository

About MCP Agent Strategy

MCP Agent Strategy is a Dify agent strategy plugin that combines MCP tool calls with common function calls. It supports configurable connections to one or more MCP servers.

Key features

  1. 01Integrate MCP tool calls
  2. 02Support the official function-call agent strategy
  3. 03Configure flexible server connections

Use cases

  1. 01Call MCP tools from a Dify agent
  2. 02Configure one or multiple MCP servers

MCP support

The plugin enables the Dify agent strategy to call tools exposed by MCP servers, with configurable URLs, headers, and timeouts.

Details

From the project README

MCP Agent Strategy

Author: hjlarry
Version: 0.0.2
Type: agent strategy
Repo&Issue: https://github.com/hjlarry/dify-plugin-mcp_agent

A Dify Agent strategy plugin that integrates MCP tool calls and common function calls capabilities.

Features

  • Full compatibility with official function call agent strategy
  • Seamless integration with MCP tool calls
  • Flexible MCP server connection options

Getting Started

agent

Basic Configuration

For single MCP server setup, simply provide the server URL:

http://localhost:8000/sse
Advanced Configuration

For multiple MCP servers with detailed connection parameters:

{
  "server_name1": {
    "url": "http://127.0.0.1:8000/sse",
    "headers": {},
    "timeout": 5,
    "sse_read_timeout": 300
  },
  "server_name2": {
    "url": "http://127.0.0.1:8001/sse"
  }
}

How to change MCP server from stdio to sse ?

Option 1: source code modification
if __name__ == "__main__":
    mcp.run(transport='sse')
Option 2: using the mcp-proxy
uv tool install mcp-proxy
mcp-proxy --sse-host=0.0.0.0 --sse-port=8080 uvx your-server

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.