Skillget

gemini-ollama

agent
Security Audit
Warn
Health Warn
  • License — License: Apache-2.0
  • Description — Repository has a description
  • Active repo — Last push 0 days ago
  • Low visibility — Only 5 GitHub stars
Code Pass
  • Code scan — Scanned 12 files during light audit, no dangerous patterns found
Permissions Pass
  • Permissions — No dangerous permissions requested

No AI report is available for this listing yet.

SUMMARY

A fully local AI CLI agent powered by Ollama. Rebranded from Gemini CLI as gmol.

README.md

gemini-ollama (gmol) 🦙

GitHub License
GitHub Stars

gemini-ollama (or simply gmol) is a powerful, privacy-focused AI agent CLI that brings the intelligence of Large Language Models directly to your terminal—100% locally.

Born from a fork of the original Gemini CLI, this project has been fully decoupled from Google Cloud services and rebranded to serve as the ultimate local-first AI companion, powered by Ollama.

✨ Why gemini-ollama?

  • 🔒 100% Private: Your code, logs, and prompts never leave your machine.
  • 🚀 Local Performance: No latency from API calls or cloud queues.
  • 🛠️ Developer-First: Designed for the terminal, with deep support for file reading, searching, and command execution.
  • 🧩 Extensible: Built-in support for skills, MCP servers, and custom agents.
  • ⚡ Universal Alias: Access your agent instantly with the gmol command.

🛠️ Prerequisites

Before using gmol, you must have Ollama installed and running on your system.

  1. Install Ollama: Download it from ollama.com.
  2. Pull a Model: gmol works best with models like Llama 3, Mistral, or Phi-3.
    ollama pull llama3
  3. Start the Server: Ensure the Ollama server is running (usually automatic after installation).

📦 Installation

This project is currently installed from source to ensure you have the latest local-first features.

# Clone the repository
git clone https://github.com/domedav/gemini-ollama.git
cd gemini-ollama

# Install dependencies and build
npm install
npm run build

# Link the binaries globally
npm link ./packages/cli

🚀 Getting Started

1. Launch the CLI

Simply type gmol in any directory to start a session.

gmol

2. Select Your Model

By default, gmol will attempt to pick the first available model from your Ollama instance. To explicitly select a model, use the slash command within the chat:

/model

Use the arrow keys to pick from your locally pulled models.

3. Usage Examples

Start with a specific model:

gmol -m llama3:8b

Include specific directories for context:

gmol --include-directories ./src,./docs

Run a one-off prompt (non-interactive):

gmol -p "Explain the logic in packages/core/src/utils/paths.ts"

⌨️ Command Overview

While in the chat interface, use these commands to control your agent:

  • /model - Change the active LLM.
  • /clear - Clear the current chat history.
  • /help - See a full list of available commands and keyboard shortcuts.

📜 License

Distributed under the Apache 2.0 License. See LICENSE for more information.

Built for developers who value privacy and local power.

Reviews (0)

No results found