https://github.com/sudo-burger/m0.nvim

Minimalistic LLM plugin for Neovim.
https://github.com/sudo-burger/m0.nvim

anthropic llm neovim-plugin openai

Last synced: 13 days ago
JSON representation

Minimalistic LLM plugin for Neovim.

• Host: GitHub
• URL: https://github.com/sudo-burger/m0.nvim
• Owner: sudo-burger
• Created: 2024-03-17T13:52:59.000Z (10 months ago)
• Default Branch: master
• Last Pushed: 2024-12-22T00:22:22.000Z (about 1 month ago)
• Last Synced: 2024-12-22T00:28:02.373Z (about 1 month ago)
• Topics: anthropic, llm, neovim-plugin, openai
• Language: Lua
• Homepage:
• Size: 400 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# m0.nvim

## Introduction

Yet another Neovim plugin interacting with LLMs.

* Chat and rewrite modes.

* Project scan (adds repo code to context).

  * Supports Anthropic prompt caching (OpenAI caching is implicit).

* Configure any number of LLM backends and prompts.

* Switch backends and prompts at any time.

* Minimal codebase.

Supported APIs:

* OpenAI: completions (https://platform.openai.com/docs/api-reference/making-requests)

* Anthropic: messages (https://docs.anthropic.com/claude/reference/messages_post)

  - Supports prompt caching for project scans.

## Installation

Using [lazy.nvim](https://github.com/folke/lazy.nvim)

``` lua

return {

  'sudo-burger/m0.nvim',

  dependencies = {

    'nvim-lua/plenary.nvim',

    'nvim-telescope/telescope.nvim',

  },

  opts = {

    providers = {

      ['anthropic'] = {

        api_key = ,

      },

      ['openai'] = {

        api_key = ,

      },

      ['mistral'] = {

        api_key = ,

      },

    },

    backends = {

      ['openai:gpt-4o-mini-nostream'] = {

        provider = 'openai',

        model = { name = 'gpt-4o-mini', max_completion_tokens = 128 },

        stream = false,

        temperature = 0.8,

      },

      ['anthropic:claude-3-haiku'] = {

        provider = 'anthropic',

        model = { name = 'claude-3-haiku-20240307' },

      },

      ['mistral:mistral-large-latest-stream'] = {

        provider = 'mistral',

        model = { name = 'mistral-large-latest', max_tokens = 3000 },

      },

    },

    default_backend_name = 'openai:gpt-4o-mini-nostream',

    prompts = {

      ['Helpful assistant'] = 'You are a helpful assistant.',

      ['Marilyn Monroe'] = 'Assume the persona of Marilyn Monroe.',

    },

    default_prompt_name = 'Helpful assistant',

    section_mark = '-*-*-*-*-*-*-*-*-',

  },

  keys = {

    { 'ab', '(M0 backend)', desc = 'M0 backend' },

    { 'ac', '(M0 chat)', desc = 'M0 chat', mode = { 'n', 'v' } },

    { 'ai', '(M0 info)', desc = 'M0 info' },

    { 'ap', '(M0 prompt)', desc = 'M0 prompt' },

    { 'ar', '(M0 rewrite)', desc = 'M0 rewrite', mode = { 'v' } },

    { 'as', '(M0 scan_project)', desc = 'M0 scan_project' },

  },

}

```

## Usage

- ":M0 chat" start/continue chat.

- ":M0 rewrite" rewrite selected text.

- ":M0 scan_project" include the current project as conversation context.

- ":M0 backend": select a backend.

- ":M0 prompt":  select a prompt.

* ":M0 info": show status and configuration.

## Similar Projects

- [karthink/GPTel](https://github.com/karthink/gptel)

- [CamdenClark/flyboy](https://github.com/CamdenClark/flyboy)

- [gsuuon/model.nvim](https://github.com/gsuuon/model.nvim)