simple-mcp-ollama-server 🚀

A minimal MCP server for remote Ollama access from VS Code\Rider via Continue plugin. Running on a Mini PC with Intel N150, 16GB RAM, Windows 11.

Hardware Setup 🖥️

• Device: Mini PC (Intel N150)
• Memory: 16GB RAM 🧠
• OS: Windows 11 🪟
• Purpose: Dedicated Ollama server for remote LLM access 🤖

Quick Start ⚡

1. Run Ollama (on Mini PC) 🔧

# Install and run Ollama, then pull a model
ollama pull Gamma3
ollama serve

2. Run MCP Server (on Mini PC) 🌐

pip install fastmcp ollama
python mcp-server.py

3. Continue plugin config.yaml (Rider, VS Code) ⚙️

name: Local Agent
version: 1.0.0
schema: v1
models:
  - name: Qwen3 1.7B
    provider: ollama
    model: qwen3:1.7b
    apiBase: "http://mini-pc-ip-address:11434"
  - name: Qwen2.5-Coder 1.5B
    provider: ollama
    model: qwen2.5-coder:1.5b-base
    apiBase: "http://mini-pc-ip-address:11434"
    roles:
      - autocomplete

rules:
  - name: Custom rule
    rule: Some custom rules text

mcpServers:
  - name: My MCP server
    type: streamable-http
    url: http://mini-pc-ip-address:8000/mcp
    schema: 0.1.0

# Context providers define what information the assistant can access 📁
context:
  - provider: code
  - provider: docs
  - provider: diff
  - provider: terminal
  - provider: problems
  - provider: folder
  - provider: codebase
  - provider: mcp

Note: 📍 Replace mini-pc-ip-address with your Mini PC's local IP address (e.g., 192.168.1.100). Ensure both devices are on the same network 🌐.