serve
Start Ollama server
LABSCO SUMMARY
Dropping to a terminal to pull a model, then again to try it.
Inference and library management are the same tool list, so trying a model you do not have yet is pull then run rather than a context switch. Most tools shell out to the ollama command and the rest call the local API at OLLAMA_HOST, so the server has to run on the machine where Ollama is installed.
What it is
An Ollama client with ten tools: prompts and OpenAI-style chat completions against local models, management of the local model library, and a tool that starts the Ollama server.
What you get
- OpenAI-compatible chat completion with system, user and assistant messages and a temperature setting, and a single-prompt run alongside it
- A tool that starts the Ollama server itself
- The local library managed in place: list, pull, push, copy and remove models
- New models created from a Modelfile
- The details of a named model, as ollama show reports them
Requirements
An Ollama installation, with the ollama command on the same machine, and Node.js with pnpm to build the server from source.
Setup effort
Build from source
USE CASES
01
Running a prompt against a model on your own machine
A single prompt to a named model for something short, or a conversation with system, user and assistant messages and a temperature when the task needs one.
02
Trying a model you have not downloaded yet
Pull it, check its details, and remove it again when it turns out to be the wrong size for the machine.
03
Making a variant with its own system prompt
Build a model from a Modelfile that sets its system prompt, copy it under another name to iterate, and push the version worth keeping.
10 tools
Listed in project code.
create
Create a model from a Modelfile
show
Show information for a model
run
Run a model
pull
Pull a model from a registry
push
Push a model to a registry
list
List models
cp
Copy a model
rm
Remove a model
chat_completion
OpenAI-compatible chat completion API
BEFORE YOU INSTALL
What to know first
Tools you get
10 tools registered in the project's code.
Requires
- Ollama installed on your system
- Node.js and npm/pnpm
Where it runs
The documented client runs the server on your machine.
CONTEXT COST
Light Standing cost
Rated by the size of the 10 tool definitions that ride along on every turn, not by how many there are
INSTALLATION
Connect your client
You clone and build this one yourself.
Server setup
TERMINAL
git clone https://github.com/NightTrek/Ollama-mcp
cd Ollama-mcp
pnpm install
pnpm run buildMCP Clients
Claude Desktop
CLAUDE DESKTOP · JSON
{
"mcpServers": {
"ollama": {
"command": "node",
"args": [
"<YOUR_PATH_TO_OLLAMA_MCP>/build/index.js"
],
"env": {
"OLLAMA_HOST": "http://127.0.0.1:11434"
}
}
}
}
