awesome-rea
Public, source-backed REA (Reverse Engineer Anything / rea-agents) case studies: games, applications, hardware and tool integrations. English-first, with 12 complete language editions.
Install / Use
claude mcp add git-sanbu -- npx -y github:git-sanbu/awesome-reaIf the server publishes to npm under a different name, use that package instead — check the repo README.
MCP Server
Model Context Protocol server
Quality Score
Category
OtherSupported Platforms
Tags
Our assessment of awesome-rea
awesome-rea scores 77/100 on our quality scale, 228th of 255 Other skills we index.
Its MCP Server is 13 KB long, well organised into 10 sections and no code examples: a thorough specification that gives an agent plenty to work with.
It has 10 GitHub stars, so there is little community track record yet; judge it on its content.
Maintenance, license and trust
- The repository was last updated today, so awesome-rea is actively maintained.
- It is released under the MIT license, a permissive license that allows use, modification and commercial use with attribution.
- Its trust signals score 97/100, with no cautions. These come from repository metadata, not a code audit — read the skill file before letting an agent act on it.
awesome-rea compared with similar skills
All 4 of these similar skills score higher than awesome-rea; compare them before choosing.
| Skill | Score | Stars | Updated | Format |
|---|---|---|---|---|
| awesome-rea (this skill)by git-sanbu | 77 | 10 | today | MCP Server |
| Agent-Reachby Panniantong | 100 | 100.5k | 1d ago | CLAUDE.md |
| headroomby headroomlabs-ai | 100 | 77.0k | today | CLAUDE.md |
| CowAgentby zhayujie | 100 | 47.5k | today | CLAUDE.md |
| Scraplingby D4Vinci | 100 | 89.0k | today | MCP Server |
Frequently asked questions
- How do I install awesome-rea?
- Run
claude mcp add git-sanbu -- npx -y github:git-sanbu/awesome-rea. The install tabs above show the steps for each supported agent. - Which AI agents does awesome-rea work with?
- It is written for Claude Code, Claude Desktop and OpenAI Codex, as a MCP Server file. Other agents that read the same format can often use it too.
- Is awesome-rea safe to use?
- It is MIT-licensed and scores 97/100 on trust signals. Skills are instructions an agent will follow, so read the file before installing it and do not approve commands you do not understand.
- Is awesome-rea still maintained?
- The repository was last updated today, so awesome-rea is actively maintained.
Skill content
View source on GitHubAwesome REA — Reverse Engineer Anything
English · 简体中文 · 繁體中文 · 日本語 · 한국어 · Español · Français · Deutsch · Português (Brasil) · Русский · العربية · हिन्दी
Understand software with REA. What will you build?
From legacy game ports to a small color picker, explore public projects using Reverse Engineer Anything. See what was built, how REA contributed, and what remains unverified.
Explore and star the collection on GitHub · Search software, purpose or keywords · Submit a case
34 public cases — 5 official showcases and 29 community projects or author experiments. Updated: 2026-10-21. Each case retains its own source-review date.
Sources reviewed; author tests not independently rerun.
CHANGELOG · Localization (English)
Popular projects
Snapshot: 2026-10-21. Source-reviewed catalog projects with at least 100 GitHub Stars, sorted by stars. Stars indicate attention, not independent reproduction or production readiness. Snapshot sources.
| Project / demo | GitHub Stars | Process and limitations | | --- | ---: | --- | | MiaAI-Lab/sparkDash | 738 | sparkDash: integrating REA analysis into a multi-device dashboard | | cranebeaker/coop-anything | 721 | Co-op Anything: investigating campaign multiplayer entry points | | michaelshimeles/terminal-rs | 259 | terminal-rs: a market terminal informed by TradingView interactions |
Official showcases
| Project / demo | The question | Primary source | | --- | --- | --- | | DX-Ball: reconstructing stereo sound panning | How does an on-screen position become a sound-pan value? | Primary source | | Notion: tracing the Electron rich-text clipboard bridge | How does a web editor write rich text through the desktop clipboard? | Primary source | | TH04: recovering PC-98 bullet-ring angles | How are bullets evenly distributed around a ring and aimed toward the player? | Primary source | | Aegis: finding the Android TOTP algorithm | How is a six-digit login code calculated every 30 seconds? | Primary source | | DownUnderCTF: recovering flag-checking equations | How can input validation be recovered without symbols or source code? | Primary source |
Games
| Project / demo | The question | Primary source | | --- | --- | --- | | Desktop Games: porting a desktop toy to macOS and the browser | How can the Peter interpreter, program tree, interactions and assets be understood and ported? | Primary source | | Ski or Die: rebuilding Snowball War in Rust | How can throwing, collision, spawning and scoring rules be recovered from a DOS game? | Primary source | | OpenBattlefield2: function-level Evidence for a Battlefield 2 client | How can crosshair, HUD and binary-behavior research inform a traceable engine implementation? | Primary source | | GunZRs: identifying a packing boundary before rebuilding from game data | How can reconstruction continue when a packed executable prevents static recovery of game logic? | Primary source | | Co-op Anything: investigating campaign multiplayer entry points | How can a campaign retain its scripts and sides while letting a second player join? | Primary source |
Applications
| Project / demo | The question | Primary source | | --- | --- | --- | | Sketch (formerly paper-rs): a Paper-inspired Rust infinite canvas | How can canvas, layout and editor interactions inform an independent design tool? | Primary source | | rec-rs: an OBS-inspired native macOS recorder | How can a focused screen, camera and audio recorder be built? | Primary source | | terminal-rs: a market terminal informed by TradingView interactions | How can chart-interaction research inform an independent Rust market interface? | Primary source | | Wispr Flow: a speech-input teardown and a small local prototype | How are recording, transcription, context and pasting connected? | Primary source | | hexgrab: investigating the PowerToys color picker | How do global hotkeys, pixel sampling and the clipboard work together? | Primary source | | AdamTerm: investigating Ghostty’s shell and building a terminal | How are the terminal core, native windows, menus and services divided? | Primary source | | Ledger Reborn: a local ledger informed by QuickBooks research | How can desktop accounting workflows inform an independent local app? | Primary source | | DiskLens: checking disk-scanning and treemap algorithms in a released binary | How can disk scanning, treemaps and duplicate detection be checked before building a new interface? | Primary source | | Hyper3D: from captured web bundles to an independent 3D prototype | How can interface and API observations inform a prototype that exports 3D models? | Primary source |
Hardware and compatibility
| Project / demo | The question | Primary source | | --- | --- | --- | | VSL-DSP: applying vendor-library findings to audio DSP tools | How does the vendor controller discover devices, register parameters and represent gain? | Primary source | | Virtual Desktop Streamer: a Linux port through a replacement platform layer | How can capture, encoding, audio and input move to Linux while retaining the protocol library? | Primary source | | Pixel Watch: an eSIM handoff helper informed by companion-APK analysis | How can a separate helper find and open the companion app’s existing watch-eSIM download flow? | Primary source | | PicoPrint: connecting a mobile thermal printer to Windows printing | When a driver connects but does not print, how can handshake differences be found? | Primary source | | V5VPN: protocol research for a scoped portal and tunnel client | How can authentication, resource discovery and connection research inform a scoped independent client? | Primary source |
Analysis and evaluation
| Project / demo | The question | Primary source | | --- | --- | --- | | OpenBSD netcat: static analysis of a shipped ELF | Which network, proxy and security-related implementations are in the shipped netcat binary? | Primary source | | REA on REA: investigating the tool and testing predictions | How does REA choose a native-analysis provider and detect an unavailable provider? | Primary source | | rea-groundtruth: checking analysis blind spots against known source | Does a report of zero IPC calls or windows reflect the app or an analysis gap? | Primary source | | rea-lab: measuring REA’s contribution to Windows reverse engineering | How does adding REA affect an agent’s success rate, time and cost? | Primary source | | bzip2: checking stripped-binary findings against a symbolized build | How can a function be found from an output string and its call relationships checked? | Primary source | | A local 27B model: a smaller MCP tool list enables function analysis | How can REA tool descriptions and analysis fit into a local model’s 32K context? | Primary source | | Open-source REA Lab: pinning an npm artifact and analyzing its entry slice | How can a release artifact be identified before answering a bounded entry-file question? | Primary source |
Learning and reproduction
| Project / demo | The question | Primary source | | --- | --- | --- | | rea-playbook: turning Electron Evidence into a report and reconstruction | How can large Evidence JSON become readable facts and a checked reconstruction? | [Primary source](https://github.com/ji
Truncated for display — read the full file on GitHub.
Related Skills
Agent-Reach
100.5kGive your AI agent eyes to see the entire internet. Read & search Twitter, Reddit, YouTube, GitHub, Bilibili, XiaoHongShu — one CLI, zero API fees.
headroom
77.0kCompress tool outputs, logs, files, and RAG chunks before they reach the LLM. 20% fewer tokens for coding agents, 60-95% fewer tokens for JSON, same answers. Library, proxy, MCP server.
CowAgent
47.5kOpen-source personal AI assistant & Agent Harness. Plans tasks, runs tools and skills, self-evolves with memory and knowledge. Multi-agent, multi-model, multi-channel. Lightweight, extensible, one-line install.
Scrapling
89.0k🕷️ A Web Scraping framework that handles everything from a single request to a full-scale crawl! Don't be shy, join here: https://discord.gg/EMgGbDceNQ and follow here for daily tips and tricks: https://x.com/Scrapling_dev
Languages
Trust signals
From repository metadata: license, adoption, age and documentation. Not a code audit — see the Safety scan above for what the skill file itself contains.
