McGravity
@tigranbs
About McGravity
Fast TUI that orchestrates AI coding tools (Claude Code, Codex, Gemini) in a plan→execute→review loop. Breaks work into atomic tasks for easier verification and course-correction.
Config
No standard config provided
This server doesn't expose a parseable MCP config block in its README. See the repository for install instructions.
RepositoryTools
No tools detected
We auto-extract tools from the README. The maintainer can list them under a ## Tools heading to populate this section.
Overview
What is McGravity?
McGravity is a fast, opinionated terminal UI (TUI) for AI-assisted coding. It orchestrates AI coding tools (Claude Code, Codex, Gemini) in a tight feedback loop: plan small tasks, execute one at a time, review, and replan. The tool is built for developers who want controlled, iterative AI assistance rather than dumping a large prompt.
How to use McGravity?
Install McGravity globally via npm (npm install -g mcgravity), then install at least one supported AI CLI tool (Claude Code, Codex, or Gemini CLI). Navigate to your project directory and run mcgravity. On first run it detects available AI tools and asks which to use for planning and execution. Type a task description and press Enter to start the flow. Settings are saved automatically to .mcgravity/settings.json.
Key features of McGravity
- Breaks work into atomic, verifiable tasks.
- Supports multiple AI backends: Claude Code, Codex, Gemini.
- File mentions with fuzzy search via
@syntax. - Slash commands for settings, clear, and exit.
- Runs entirely locally; no code or keys transmitted.
- Configurable max iterations and enter behavior.
Use cases of McGravity
- Break a large feature request into small, sequential tasks.
- Review and course-correct after each AI-generated change.
- Build a new endpoint by planning, implementing, and testing incrementally.
- Fix bugs with controlled, iterative AI suggestions.
FAQ from McGravity
Which AI tools does McGravity support?
McGravity supports Claude Code, Codex (OpenAI), and Gemini CLI (Google). You need at least one installed.
How do I install McGravity?
Install it with npm install -g mcgravity, or build from source using cargo install --path . after cloning the repository.
Does McGravity send my code anywhere?
No. McGravity runs AI CLI tools locally on your machine and never collects, stores, or transmits your code or API keys.
Where are settings stored?
Settings are saved automatically to .mcgravity/settings.json in your project directory.
How do I insert a newline in the task input?
Use Ctrl+J (works on all terminals), Shift+Enter, or a backslash escape (\ then Enter). You can also set "Enter Behavior" to "Newline" in settings.
Frequently asked questions
Which AI tools does McGravity support?
McGravity supports Claude Code, Codex (OpenAI), and Gemini CLI (Google). You need at least one installed.
How do I install McGravity?
Install it with `npm install -g mcgravity`, or build from source using `cargo install --path .` after cloning the repository.
Does McGravity send my code anywhere?
No. McGravity runs AI CLI tools locally on your machine and never collects, stores, or transmits your code or API keys.
Where are settings stored?
Settings are saved automatically to `.mcgravity/settings.json` in your project directory.
How do I insert a newline in the task input?
Use `Ctrl+J` (works on all terminals), `Shift+Enter`, or a backslash escape (`\` then `Enter`). You can also set "Enter Behavior" to "Newline" in settings.
Basic information
More AI & Agents MCP servers
bsv.cx
Andy RoweTimestamp and verify evidence on-chain. Let your agent prove what it saw and when. bsv.cx gives an AI agent verifiable memory. Timestamp a hash of any output, archive a web page as it looked, or fetch a URL with a recei
LinkedIn MCP Server
stickerdanielOpen-source MCP server for LinkedIn. Give Claude and any MCP-compatible AI agent access to profiles, companies, jobs, and messages.
欢迎来到 智言平台
Shy2593666979AgentChat 是一个基于 LLM 的智能体交流平台，内置默认 Agent 并支持用户自定义 Agent。通过多轮对话和任务协作，Agent 可以理解并协助完成复杂任务。项目集成 LangChain、Function Call、MCP 协议、RAG、Memory、HITL、Skill、Milvus 和 ElasticSearch 等技术，实现高效的知识检索与工具调用，使用 FastAPI 构建高性能后端服务。
Shell and Coding agent for Claude and other mcp clients
rusiaamanShell and coding agent on mcp clients
Model Context Protocol for Unreal Engine
chongdashuEnable AI assistant clients like Cursor, Windsurf and Claude Desktop to control Unreal Engine through natural language using the Model Context Protocol (MCP).
Comments