avante.nvim  by yetone

Neovim plugin emulating Cursor AI IDE for AI-driven code assistance

created 11 months ago
15,381 stars

Top 3.3% on sourcepulse

GitHubView on GitHub
Project Summary

avante.nvim aims to bring Cursor AI IDE-like code assistance to Neovim. It allows users to query AI about their code, receive suggestions, and apply them directly, enhancing developer productivity. The target audience includes Neovim users seeking advanced AI-powered code editing capabilities.

How It Works

avante.nvim integrates with various LLM providers (OpenAI, Claude, Ollama, etc.) to process code queries. It leverages Neovim's Treesitter for code understanding and offers features like AI-driven code suggestions, diff application, and a sidebar for managing AI interactions. The plugin supports both "agentic" and "legacy" modes for AI code generation.

Quick Start & Requirements

  • Installation: Recommended via lazy.nvim (see README for other package managers).
  • Prerequisites: Neovim 0.10.1 or later. API keys for chosen AI providers (e.g., ANTHROPIC_API_KEY, OPENAI_API_KEY) are recommended as environment variables.
  • Setup: Minimal setup involves adding the plugin and calling require('avante').setup().
  • Docs: https://github.com/yetone/avante.nvim

Highlighted Details

  • Supports multiple AI providers including OpenAI, Claude, Ollama, and custom providers.
  • Features a sidebar for AI chat, diff viewing, and file selection.
  • Includes tools for web search, Git operations, and file manipulation.
  • Offers customizable prompts and custom tool integration.

Maintenance & Community

The project is actively developed with ongoing iterations and feature additions. Community contributions are welcomed. Links to potential community channels are not explicitly provided in the README.

Licensing & Compatibility

  • License: Apache 2.0 License.
  • Compatibility: Compatible with commercial use and closed-source linking.

Limitations & Caveats

The plugin is in active development, with functionality and interface subject to significant changes. Some features, like auto-suggestions, are experimental. The README notes that certain models may have suboptimal performance or tool usage, recommending specific models or modes for better results.

Health Check
Last commit

13 hours ago

Responsiveness

1 day

Pull Requests (30d)
109
Issues (30d)
52
Star History
2,402 stars in the last 90 days

Explore Similar Projects

Feedback? Help us improve.