localise-opencode

Auto-detect and configure local LLMs for opencode.

Version: 1.0.0

Installation

pip install localise-opencode

Or install from source:

git clone https://github.com/praneel/localise-opencode.git
cd localise-opencode
pip install -e .

Prerequisites

• LM Studio running with a model loaded
• opencode installed

Usage

Detect running servers

localise-opencode detect

List available models

localise-opencode list

Add a model to opencode

# Interactive mode - shows available models
localise-opencode add

# Add specific model
localise-opencode add llama-3.1-8b-instruct

Check configuration

localise-opencode status

Remove a model

# Interactive mode
localise-opencode remove -i

# Remove specific model
localise-opencode remove llama-3.1-8b-instruct

Configuration

The tool modifies ~/.config/opencode/opencode.jsonc to add your local LLM provider.

Example configuration added:

{
  "provider": {
    "lmstudio": {
      "name": "LM Studio",
      "base_url": "http://localhost:1234/v1",
      "api_key": "lm-studio",
      "models": {
        "llama-3.1-8b-instruct": {}
      }
    }
  }
}

Development

# Install dev dependencies
pip install -e ".[dev]"

# Run linter
ruff check src/

# Run tests
localise-opencode test

License

MIT