MCP.so
登录

Ollama-MCP Bridge WebUI

@Rkm1999

关于 Ollama-MCP Bridge WebUI

A web interface connecting local Ollama LLMs to Model Context Protocol (MCP) servers. Enables open-source models to use file operations, web search, and reasoning tools similar to commercial AI assistants - all running privately on your own hardware.

基本信息

分类

AI 与智能体

许可证

MIT

运行时

node

传输方式

stdio

发布者

Rkm1999

配置

暂无标准配置

该服务器的 README 中没有可解析的 MCP 配置块,请前往代码仓库查看安装说明。

代码仓库

工具

未检测到工具

工具是从 README 中自动提取的。维护者可以在 ## Tools 标题下列出工具,即可填充这部分内容。

概览

What is Ollama-MCP Bridge WebUI?

A TypeScript implementation that connects local LLMs (via Ollama) to Model Context Protocol (MCP) servers with a web interface. It allows open-source models to use the same tools and capabilities as Claude, enabling powerful local AI assistants that run entirely on your own hardware.

How to use Ollama-MCP Bridge WebUI?

Run the included installation script (./install.bat) for automatic setup, then add your API keys to the .env file. Start the bridge with ./start.bat, and open http://localhost:8080 in your browser to access the web interface.

Key features of Ollama-MCP Bridge WebUI

  • Multi-MCP Integration: Connect multiple MCP servers simultaneously.
  • Tool Detection: Automatically selects the right tool for queries.
  • Web Interface: Clean UI with collapsible tool descriptions.
  • Comprehensive Toolset: Filesystem, web search, and reasoning capabilities.
  • Local Execution: Runs entirely on your hardware via Ollama.
  • Open-source Models: Uses models like Qwen via Ollama.

Use cases of Ollama-MCP Bridge WebUI

  • Querying local files using natural language through the filesystem server.
  • Performing web searches with the Brave Search integration.
  • Solving complex problems step-by-step with sequential thinking.
  • Running a private AI assistant without relying on cloud services.
  • Using open-source models with Claude-like tool capabilities.

FAQ from Ollama-MCP Bridge WebUI

What dependencies are required?

Node.js and Ollama are checked during automatic installation. The bridge also requires the Qwen model and npm dependencies installed manually or via the script.

Which LLM model is used by default?

The default model is qwen2.5-coder:7b-instruct-q4_K_M, as shown in the bridge_config.json example.

How do I configure API keys?

Add your API keys to the .env file. The $VARIABLE_NAME references in the configuration will be replaced with actual values from the .env file.

What MCP servers are bundled?

The bridge includes Filesystem, Brave Search, and Sequential Thinking servers by default, configured in bridge_config.json.

How do I start the bridge?

Run ./start.bat to start the bridge with the web interface, then open http://localhost:8080 in your browser.

评论

AI 与智能体 分类下的更多 MCP 服务器