光速 — Speed Of Light ⚡️

光速 (Speed Of Light) is a Neovim setup that has been carefully crafted to be feature-rich yet blazingly fast due to lazy loading as much as possible!

💻 Installation

To install, simply run the ./install.sh script or run this command.

nvim --headless "+Lazy! sync" +qa

📦 Dependencies

• fd
• gcc or clang
• git
• imagemagick
• node (with npm)
• pngpaste or xclip
• python3 (with pip and virtualenv)
• ripgrep
• tectonic

📁 File Structure

Plugins are stored and automatically sourced in lua/plugins/

~/.config/nvim
├── ftplugin
│   └── *.lua
├── init.lua
├── lazy-lock.json
├── lua
│   ├── core
│   │   ├── autocmd.lua
│   │   ├── functions.lua
│   │   ├── mappings.lua
│   │   └── options.lua
│   ├── neovide
│   │   └── init.lua
│   ├── plugins
│   │   └── *.lua
│   └── versions.lua
└── README.md

🤖 AI Features

There are 3 AI features that are enabled by default:

1. Supermaven
2. Avante
3. McpHub

Supermaven

Supermaven should already be enabled by default, but to use the pro version, run :SupermavenUsePro and follow the instructions.

Avante

Currently, Avante is configured to use the following models from the following providers:

Model Name	Provider
gemini-2.5-flash	Google AI
mistral-large-latest	Mistral AI
deepseek-r1	OpenRouter
llama-3.3	Groq

macOS

How to store API Keys

security add-generic-password -a "Gemini API Key" -s "GEMINI_API_KEY" -w "<api_key>"
security add-generic-password -a "Mistral API Key" -s "MISTRAL_API_KEY" -w "<api_key>"
security add-generic-password -a "OpenRouter API Key" -s "OPENROUTER_API_KEY" -w "<api_key>"
security add-generic-password -a "Groq Token" -s "GROQ_API_KEY" -w "<api_key>"

Linux

How to store API Keys

printf "<api_key>" | secret-tool store --label="Gemini API Key" token GEMINI_API_KEY
printf "<api_key>" | secret-tool store --label="Mistral API Key" token MISTRAL_API_KEY
printf "<api_key>" | secret-tool store --label="OpenRouter API Key" token OPENROUTER_API_KEY
printf "<api_key>" | secret-tool store --label="Groq API Key" token GROQ_API_KEY

McpHub

Currently, McpHub is configured to use the following servers from the following authors:

MCP Server	Author/Provider
Playwright	executeautomation
Sequential Thinking	modelcontextprotocol
Context7	upstash
Neovim	Native
Docker MCP Toolkit	Docker

Configuration

{
  "mcpServers": {
    "docker-desktop-mcp-toolkit": {
      "disabled": true,
      "autoApprove": [],
      "command": "docker",
      "args": ["mcp", "gateway", "run"]
    },
    "github.com/modelcontextprotocol/servers/tree/main/src/sequentialthinking": {
      "autoApprove": ["sequentialthinking"],
      "custom_instructions": {
        "text": "Input parameters should use camelCase, not snake_case. For example, thoughts_needed should be changed to thoughtsNeeded."
      },
      "args": ["-y", "@modelcontextprotocol/server-sequential-thinking"],
      "command": "npx",
      "disabled": true
    },
    "github.com/upstash/context7-mcp": {
      "disabled": false,
      "autoApprove": ["resolve-library-id", "get-library-docs"],
      "command": "npx",
      "args": ["-y", "@upstash/context7-mcp"]
    },
    "github.com/executeautomation/mcp-playwright": {
      "disabled": true,
      "command": "npx",
      "args": ["-y", "@executeautomation/playwright-mcp-server"]
    }
  },
  "nativeMCPServers": {
    "neovim": {
      "disabled": true,
      "autoApprove": []
    },
    "mcphub": {
      "disabled": false,
      "autoApprove": ["get_current_servers", "toggle_mcp_server"]
    }
  }
}