Skip to content

installation.md

Latest commit

 

History

History
78 lines (56 loc) · 2.39 KB

installation.md

File metadata and controls

78 lines (56 loc) · 2.39 KB

Installation & Setup Guide

Mimir is distributed as a single static binary. You can install it using several methods depending on your environment and preference.

1. Automated Installation (Recommended)

For Linux and macOS users, the fastest way is to use the one-line installer. This script automatically detects your OS and architecture (Intel or Apple Silicon/ARM) and installs the correct binary to /usr/local/bin.

curl -fsSL https://raw.githubusercontent.com/thuongh2/git-mimir/main/install.sh | sh

2. Manual Download

If you prefer to install manually or are on a system where the script doesn't work:

  1. Go to the GitHub Releases page.
  2. Download the tarball corresponding to your system:
    • mimir-linux-amd64.tar.gz (Standard Linux)
    • mimir-linux-arm64.tar.gz (Linux on ARM/Raspberry Pi)
    • mimir-darwin-amd64.tar.gz (Intel Mac)
    • mimir-darwin-arm64.tar.gz (Apple Silicon Mac)
  3. Extract the archive: tar -xzf mimir-*.tar.gz.
  4. Move the mimir binary to a directory in your $PATH (e.g., /usr/local/bin).

3. Go Toolchain (For Developers)

If you have Go 1.22+ installed, you can compile from source or use go install:

# Global install
go install github.com/thuongh2/git-mimir/cmd/mimir@latest

# Or clone and build manually
git clone https://github.com/thuongh2/git-mimir.git
cd git-mimir
make build

4. Post-Installation Verification

Run the following command to verify the installation:

mimir --version

5. Setting up MCP for AI Agents

Mimir is designed to work perfectly with Claude Code and other MCP-compatible clients.

Automatic Setup

Run the setup command to automatically configure your editor (VS Code, Claude Desktop, etc.):

mimir setup

Manual Configuration

If automatic setup isn't supported for your tool, add the following to your MCP configuration file (usually claude_desktop_config.json):

{
  "mcpServers": {
    "mimir": {
      "command": "mimir",
      "args": ["mcp"]
    }
  }
}

6. Required Dependencies for Full Features

  • Search & Graph: No extra dependencies.
  • Semantic Search (Vector): Requires an embedding provider. Mimir supports:
    • Ollama (Local): Recommended for privacy. Just run ollama run nomic-embed-text.
    • OpenAI: Set your OPENAI_API_KEY environment variable.