開発ツール
potpie
A context graph for the AI-native SDLC — so agents and humans share one map of the codebase.
★ 5.7k スターApache-2.0
ガイドを見る →reaは、morlutoが公開しているオープンソースの開発ツールプロジェクトです（GitHub: morluto/rea）。GitHubスターは21,368、フォークは2,305で、MITライセンスの下で公開されています。GitHubでの最終更新は2026年10月8日です。
Reverse engineer anything with agents - from app behavior down to native binaries
Choose the app, binary or code path whose behavior you need to understand.
Run the one-command setup with npx - no manual clone is needed for normal use.
Let the setup register REA's MCP server and skill with your agent, then restart the agent.
Ask your agent to analyze the target and review the evidence-backed findings it returns.
A single MCP server covers binaries, applications and runtime behavior for your agent.
Returns modules, imports, boundaries and the evidence behind each conclusion, not just guesses.
Works with established analysis tools - the project tags Ghidra and Hopper among its ecosystem - for native analysis.
The same workflows run through your agent via MCP or from the terminal via the rea command.
The README demonstrates cases like tracing Notion's Electron clipboard bridge and reconstructing game logic from a DOS binary.
You can start in demo mode before activating any license the setup prompts for.
Get the source code from GitHub and enter the project directory.
git clone https://github.com/morluto/rea.git
cd reaOne npx command installs and registers REA with your coding agent.
npx rea-agents setupIf you use npx, refresh your agent registrations and skill with the latest package.
npx rea-agents@latest setup # see README Update sectionIt is maintained by morluto and contributors on GitHub.
REA is MIT licensed and free to use as open source; the setup may offer demo mode or license activation for parts of the product - see the README for details.
REA works through your own AI agent, so you need a Claude Code, Codex or similar agent setup with its own model access; REA itself is the analysis layer.
Compared with driving Ghidra or a disassembler by hand, REA lets the agent orchestrate the analysis and returns structured, evidence-linked findings in natural language.
Run npx rea-agents setup, restart your agent, and ask it to analyze a target - the README's showcases show the style of questions it handles.
A context graph for the AI-native SDLC — so agents and humans share one map of the codebase.
Xiaomi's terminal-native AI coding assistant where models and agents co-evolve, with project memory.
Hunt down the lowest-cost AI coding setup your repo can rely on.
Source control for agents — run coding agents side by side, track changes and query their history.
A guard that blocks dangerous git and shell commands before AI agents can execute them.
A virtual terminal and filesystem that lets AI agents work across S3, Drive, Slack, Gmail and more.
LibreAIカタログに掲載された開発ツールプロジェクトの中で最もスターの多いもの — reaが合わなかった場合に検討したい、現実的な候補リストです。
The open source coding agent
The model-definition framework behind modern AI — thousands of pretrained transformers for text, vision, and audio.
Anthropic's agentic coding tool that lives in your terminal and works across your whole codebase
Cross-platform desktop all-in-one manager for Claude Code, Codex, OpenCode and other coding agents
OpenAI's lightweight coding agent that runs locally in your terminal
カタログの中でreaとスター数が近い開発ツールプロジェクト — 前後の棚もそのまま見て回れます。
Fine-tune huge models on small hardware with LoRA and other efficient methods
Self-hosted TikTok and Douyin scraper API - no-watermark downloads, MCP and web console
Hugging Face's library for loading and processing datasets for AI models in one line
A high-performance coding agent harness in Rust - fast TUI, swarms and every provider
A desktop GUI and toolkit for Claude Code - agents, sessions, analytics and MCP
Turn any GitHub repo into an interactive architecture diagram - just swap 'hub' for 'diagram'