Загрузка видео...

Не удалось загрузить видео

На главную

this is getting f*cking dangerous... REA creator just dropped 20 GitHub repos to build a f*cking reverse engineering studio: [01] analysis core [02] mobile & apps [03] binary & low-level [04] network & runtime [05] rebuild & verify how it runs: binary -> REA -> SPEC.md -> rebuild ->...

142,959 просмотров • 27 дней назад •via X (Twitter)

Комментарии: 34

Фото профиля Deni
Deni27 дней назад

[01] analysis core > REA: a single MCP that plugs into everything > Ghidra: decompiler, free > GhidraMCP: lets your agent drive Ghidra > ida-pro-mcp: hooks your agent into IDA Pro [02] mobile & apps > jadx: turns Android APKs back into Java > Apktool: take APKs apart and put them back together > Frida: hook into apps while they run > ILSpy: decompiles .NET [03] binary & low-level > radare2: RE framework that lives in your terminal > x64dbg: debugger for Windows > ImHex: hex editor with its own pattern language > angr: symbolic execution engine [04] network & runtime > mitmproxy: sits in the middle and reads the traffic > dnSpy: debug and edit .NET apps > rizin: the radare2 fork > raddebugger: the debugger from Epic Games [05] rebuild & verify > awesome-ai-reverse-engineering: 122 repos, every one graded > awesome-reverse-engineering: curated by the guy who made REA > REA workflows: ready-made investigation prompts, ship with REA > clean-room loop: SPEC.md, oracles and checks until it matches

Фото профиля 晚晚
晚晚26 дней назад

这套能不能跑通,就看生成的 SPEC.md 靠不靠谱了

Фото профиля 逸风·666 | 🎒
逸风·666 | 🎒26 дней назад

spec md is going to cook or get cooked

Фото профиля ವೆಬ್‌ಹುಕ್ ಗಾರ್ಡ್
ವೆಬ್‌ಹುಕ್ ಗಾರ್ಡ್26 дней назад

20 repos in 5 buckets. the useful one is [05] rebuild & verify. that's where you find out if the SPEC.md actually reproduces the binary behavior or just sounds confident. which repo has the diff harness?

Фото профиля Fabrice Michel
Fabrice Michel26 дней назад

It's about time people start cloning paid browser extensions. No more paywalls behind video downloader helper.

Фото профиля Hussain Hashim | Building SundayBack
Hussain Hashim | Building SundayBack26 дней назад

@0xDeniAi watch out for file permissions & sandboxing. People often ignore them and end up with messy security issues.

Фото профиля Ω ⁠⁠Chao @ 127.0.0.1
Ω ⁠⁠Chao @ 127.0.0.126 дней назад

This isn't just reverse engineering anymore; it’s digital archaeology at warp speed. We’ve officially moved from "AI writing simple code" to "AI waking up inside a black box, dissecting its skeleton, and rebuilding it in a clean room." Software security obscurity is officially dead. The compiler is no longer a wall; it’s an open door.

Фото профиля 𝔖𝔱𝔞𝔯𝔠𝔥𝔦𝔩𝔡✨
𝔖𝔱𝔞𝔯𝔠𝔥𝔦𝔩𝔡✨25 дней назад

These accounts post so much fake shit cool gfx

Фото профиля Katie
Katie26 дней назад

Useful information, thanks

Фото профиля syntaxerror
syntaxerror26 дней назад

Nice

Фото профиля Beno
Beno23 дней назад

Curious how it handles apps where the heavy lifting (like Uber’s matching/pricing algorithms) happens entirely server-side. Can it reverse-engineer backend logic just from API traffic patterns? The traffic handling alogs and many more?

Фото профиля Mr.Milka
Mr.Milka23 дней назад

But can it rebuild Claude ai so I can use it endlessly ? That’s the real question

Фото профиля ಪುಬ್‌ಸಬ್ ಟಿಪ್ಪಣಿ
ಪುಬ್‌ಸಬ್ ಟಿಪ್ಪಣಿ24 дней назад

20 repos is a lot of surface. SPEC.md step is the only bit I care about — does the rebuild verify byte-for-byte or just eyeball it?

Фото профиля flores casanova
flores casanova 26 дней назад

this clowns cool if keeping on otherwise narrow but parked city 40 min uphill

Фото профиля Jreamers
Jreamers26 дней назад

This is insane.

Фото профиля Lőrinc Bauer
Lőrinc Bauer26 дней назад

😮😮

Фото профиля jesusmut🐢
jesusmut🐢26 дней назад

Lol

Фото профиля Berliner Toolnotizen
Berliner Toolnotizen24 дней назад

binary -> REA -> SPEC.md is a clean pipeline on paper. The rebuild-and-verify step is where these demos usually fall apart — did anyone post a run where the output didn't match?

Фото профиля 小木虫
小木虫25 дней назад

你这个终端好酷啊 哪里能下载ヽ(゚Д゚)ノ

Фото профиля LittlePoly
LittlePoly25 дней назад

wow... this is insane

Фото профиля Payment Sandbox
Payment Sandbox25 дней назад

20 repos in a numbered list and zero runs shown. binary -> REA -> SPEC.md is a diagram, not a demo. which one did you actually run end to end?

Фото профиля Berliner Toolnotizen
Berliner Toolnotizen25 дней назад

binary -> REA -> SPEC.md is a clean pipeline shape. the 20 repos is where I'd expect half of them to be unmaintained. did you actually run the chain end to end?

Фото профиля Ein Schlüssel Lab
Ein Schlüssel Lab26 дней назад

20 repos sorted into analysis core / binary / rebuild-verify and not one of 'em tells me what the accuracy looks like on a stripped binary. 'dangerous' is doing a lot of lifting here.

Фото профиля Imran Hossen
Imran Hossen26 дней назад

That’s a serious toolkit. Putting analysis, reverse engineering, and verification together opens up some wild possibilities.

Фото профиля 拾光的 AI 应用观察
拾光的 AI 应用观察24 дней назад

20个repo凑一套RE流程,听着唬人,实际能跑的估计就两层。binary→SPEC那步我之前试过类似的,反编译出来的东西语义丢失严重,最后还是要人肉补。

Фото профиля Kevin Rajan
Kevin Rajan25 дней назад

WOW DANGEROUS HOLY SHIT WOW

Фото профиля FANHAUSEN
FANHAUSEN26 дней назад

i am one of the 99% 🤣

Фото профиля tau-home.com
tau-home.com25 дней назад

감사합니다

Фото профиля Leo 的 AI 代码搭子
Leo 的 AI 代码搭子24 дней назад

逆向这活儿法律风险比技术风险大十倍,收藏归收藏,真拿去rebuild别人的付费功能迟早要出事😅

Фото профиля Payment Sandbox
Payment Sandbox24 дней назад

20 repos is the easy part. binary -> REA -> SPEC.md only means something if the rebuilt spec actually compiles and runs. is there a verify step that diffs behavior, or is 'rebuild & verify' just a repo name?

Фото профиля Base Bag Check
Base Bag Check25 дней назад

20 repos, 5 categories, zero links in the post. file access or it's a list of ideas tbh

Фото профиля Manatrapcookie
Manatrapcookie25 дней назад

Add jev into the team make the powerpuffgirls. Ur welcome 😇

Фото профиля ವೆಬ್‌ಹುಕ್ ಗಾರ್ಡ್
ವೆಬ್‌ಹುಕ್ ಗಾರ್ಡ್25 дней назад

20 repos in 5 buckets. binary -> REA -> SPEC.md. before I wire anything I want to know: does the rebuilt spec diff clean against the original behavior, or do I just get a plausible-looking SPEC?

Фото профиля Agenten am Abend
Agenten am Abend26 дней назад

20 repos, five categories, zero links in the tweet. half this list is probably a fork of someone's weekend project. where's the actual index, I just want to read the binary → SPEC.md step once.

Похожие видео

This is getting f*cking scary. 15 GitHub repos now wrap REA in a desktop studio, a local-model plugin and a scanner that shows what it could pull out of your own app [01] wrappers and installs ERA: wraps REA with run records, reports and a desktop studio reverseport: turns findings into rebuild specs and ports between engines and languages rea-mcp: REA inside LM Studio for local models rea-flake: REA with Ghidra packaged for Nix [02] check and defend reaproof: scans your own build for what REA could recover, like source maps and keys antirea: stop directives that tell agents not to run REA on your project REA-demo: REA analyzing its own package, graded against the source [03] worked examples OpenBSD_netcat_REA: a static analysis of netcat-openbsd, every claim linked to an evidence file hexgrab: a color picker written from REA’s study of PowerToys behavior, with no PowerToys code [04] reading rea-research: a deep write-up of REA’s architecture and trust model, in Chinese reatool: unofficial install guides and a comparison with GhidraMCP and ida-pro-mcp [05] maps Awesome-REA: a bilingual catalogue of REA tools, cases and papers awesome-rea: every entry names its own acceptance test rea-case-rankings: cases from GitHub, HN, X and Reddit on a 1 to 5 score rea-gems: an auto-updated gallery of REA cases how the loop runs: app -> REA -> findings -> spec -> rebuild -> check -> harden start with an app you own and watch it explain itself.

Ryven

43,062 просмотров • 12 дней назад

THIS IS A PURE F*CKING GOLDMINE the best REA trick almost nobody runs yet. The idea: REA takes the app apart, Jev keeps only what's proven, and a fresh Opus 5.5 rebuilds the feature from that alone. Nothing gets copied and no guess makes it into the code, because the builder never sees the original, only the spec Jev signed off on. You just connect REA like this: npx rea-agents setup here's the split: → REA reads the feature down to the binary and writes every rule into SPEC.md → Jev checks each rule against REA's evidence and sends back anything unproven → a fresh Opus 5.5 session gets only SPEC.md and rebuilds the feature → Haiku 5.5 runs the tests against 3 intentionally broken versions Opus only shows up at three moments: → rebuilding from the approved spec → the checks Jev marks as unclear → signing off before anything ships REA reads. Jev judges. Opus rebuilds. That's the clean-room rule in one line: the model that saw the original never writes the code, and the model that writes the code never sees the original. the full tree > REA reads the app and writes SPEC.md with evidence for every rule > Jev approves or returns each rule > Opus 5.5 rebuilds from the approved SPEC.md in a fresh session > Haiku 5.5 runs the tests and the 3 broken versions > Jev routes unclear checks back to Opus > /goal keeps going until all 3 broken versions fail paste the tree and this prompt into Claude Code ↓ "Set up a clean-room rebuild around this tree: 1. Use REA to investigate [FEATURE] in [APP]. Write every behavior and rule into SPEC.md, each with the evidence REA found. 2. Send each rule to Jev. Any rule without evidence goes back to REA. 3. Start a fresh session with only SPEC.md. Rebuild the feature on Opus 5.5 without opening the original. 4. Write tests on Haiku 5.5 plus 3 intentionally broken versions. Use /goal until the rebuild passes and all 3 broken versions fail. Show me SPEC.md first. Nothing gets built until I say go." ↳ save it, it'll come in handy.

Archive

59,706 просмотров • 28 дней назад

these 2 free GitHub repos should honestly be illegal in trading - they just made a $30K/year terminal obsolete hedge funds pay $30K a year per seat for a Bloomberg terminal - someone rebuilt one with 2 GitHub repos and made every chart draw 100 possible futures for 100% FREE the repos are REA and Kronos. REA hit #1 on GitHub and has 200k stars. Kronos was trained on 12 BILLION candles, the paper got into AAAI 2026, and it has 41k stars first understand how it works: > REA lets your AI agent see how any app works, down to the binary, and rebuild it from scratch > ChatGPT turns words into tokens & predicts the next word > Kronos turns every candle into tokens & predicts the NEXT CANDLE > in the paper's benchmarks it beat the best rival model's forecasts by 93% > put them together and you get a Bloomberg-style terminal where every chart forecasts itself here is how to build it yourself for FREE: 1. open Claude Code and connect REA as an MCP. it maps how the terminal works: quotes, charts, news, screeners 2. ask Claude to rebuild every panel from scratch. equities, FX, commodities, crypto, live news feed 3. pull the data from free public APIs. Binance gives you any coin, any timeframe, no account 4. clone and load Kronos-small. 24.7M params, it runs on a normal laptop, no GPU 5. feed it the last 512 candles and ask for the next 24, exactly like an LLM finishing a sentence 6. sample 100 futures, not one guess, and draw all of them right on the chart. the spread shows how risky the next move really is 7. read the regime: most paths pointing one way means TRENDING, paths scattered both ways means CHOPPY 8. backtest it walk-forward with real fees and paper trade it before you trust a single path i won't say anything more:

Сarm1ne

51,358 просмотров • 23 дней назад