@swem_null

N/A

Joined August 2021
  • Tweets273
  • Following391
  • Followers14
3Media
This was super easy to do, Now my tv running super smooth same as new one - Added my own lightweight UI - custom wallpaper slideshow - switch between hindi/english for my parents - github contribution chart and bunch of other feature i have in mind to try on my tv
Gathin (Ø,G)
@gathintOct 17
Vibe coded my parents' TV this afternoon! It was showing a sponsored ad before you could even get to Netflix. Gave Claude Code the model number, it found the TV on wifi, removed 25 Xiaomi packages and wrote a new launcher. Looks like an Apple TV now lol
8
14
3
227
67,245
swem retweeted
codila
@0xCodilaOct 17
F*ck YES... I made prompt that reverse-engineer anything without BAN's and rejects the shift: REA + free model pull out all app's binary code → Opus 5.5 builds everything else the way YOU want here's 3 step guide: step 1 → install OpenCode and choose DeepSeek V4.1 or GLM-5.3 flash (both free right now) step 2 → drop prompt image + your app/file/URL into chat and tell it what to reverse (it start write repo) step 3 → open repo with full app code in Claude Opus 5.5 and say something like: "I've shared repo for my project I stopped working on. I want to pick it back up and make it even better with you! Help me continue development" the result: you can use mechanics of any website/app in your own projects... Paste it now. Then read article below to master REA:
codila
@0xCodilaOct 12
REA is most dangerous thing on the internet right now... It lets your AI agent reverse engineer any app, down to the binary If you use it correctly, you can rebuild any feature you pay for, for FREE In this article, I show you how:

Reverse Engineering: How to Get Anything for FREE With Most Dangerous Repo & Opus 5.5 (Full Course)

Article · @0xCodila
60
525
18
5,226
189,903
ThreatBook launched RedAgents, an AI-powered pentest aggregator enabling SecOps to run multiple autonomous agents (Artex Pro, Cyber(YanJi), Yanxun, CyberStrikeAI) in parallel to uncover more vulnerabilities with unified results. threatbook.io/press-releases…

ThreatBook Launches AI-Powered Pentest Aggregator, RedAgents

ThreatBook introduces RedAgents, an AI-powered pentest aggregator that enhances security operations by using multiple agents to identify vulnerabilities more effectively.

threatbook.io
19
93
5,232
swem retweeted
NOMVEL
@NomvelHQOct 15
Claude Sonnet 4.6 Free API 3M free tokens every new account apmix.ai/models/claude-sonne…
2
24
177
9,123
claude + REA can rebuild software features without copying a single line of code opus 5.5 reads what REA finds inside a program, then rebuilds the feature in a separate clean-room setup. here's how to do it: 1. run npx rea-agents setup to connect REA to Claude Code 2. pick a feature and let REA figure out how it works. save the behavior and rules in a spec file 3. start a fresh Claude session with only that spec. rebuild the feature without touching the original code 4. use /goal to run the tests and check 3 intentionally broken versions. all 3 must fail one project recovered 697 functions in just 12 days, with 696 exact matches. the real trick is building tests that catch mistakes before they become part of the finished software.
NO1ennn
@N01ennnOct 14
REA is pure f*cking treasure right now it lets your AI agent reverse engineer any app, down to the binary Opus 5.5 reads how a feature works, then rebuilds its own version from scratch but one wrong number can pass every check in this article :

How to Turn Opus 5.5 Into a Reverse Engineer: The Full Guide to REA, Oracles and Clean-Room Loops

Article · @N01ennn
115
300
12
1,828
138,979
AI in Web3 Security: Build Your First Bug-Hunting Agent Blog: 0xasen.xyz/build-your-first-… Author: @asen_sec
1
24
167
7,315
AI 에이전트를 위한 다크 웹 및 위협 인텔리전스. HIBP, ThreatFox, 랜섬웨어 추적, Tor .onion 접속, 블록체인 정보, 익스플로잇 검색, 스틸러 로그, 멀웨어 분석 등 모든 기능이 단일 MCP 서버에 통합되어 있습니다. 이제 AI 에이전트는 16개의 브라우저 탭을 열고 수동으로 상관관계를 분석할 필요 없이, 필요에 따라 다크 웹에 대한 모든 정보를 즉시 얻을 수 있습니다 .github.com/badchars/darknet-…
12
100
443
20,999
swem retweeted
Deni
@0xDeniAiOct 13
REA + Opus 5.5 is f*cking dangerous… I'm honestly not sure this should be public. 10 repos that let an AI agent take apart any app like a mechanic takes apart an engine. send to your Claude: •REA (171K stars) - one MCP for RE: binaries, Electron, .NET, Android, websites. github.com/morluto/rea • Chrome DevTools MCP (53K stars) - agent drives a live Chrome: network, console, JS. Maps how any web app works. github.com/ChromeDevTools/ch… • JADX (50K stars) - Android Dex → Java decompiler. GUI + CLI. github.com/skylot/jadx • x64dbg (49K stars) - Windows debugger built for RE and malware analysis. github.com/x64dbg/x64dbg • Apktool (26K stars) - unpacks and rebuilds APKs: manifest, resources, smali. github.com/iBotPeaches/Apkto… • radare2 (25K stars) - UNIX-style RE toolkit. Has its own MCP server (r2mcp). github.com/radareorg/radare2 • Cutter (20K stars) - free GUI RE platform powered by rizin. github.com/rizinorg/cutter • IDA Pro MCP (12K stars) - connects IDA Pro to your agent over MCP. github.com/mrexodia/ida-pro-… • dnSpy (11K stars) - .NET debugger + assembly editor. Patch and rerun. github.com/dnSpyEx/dnSpy • angr (9K stars) - Python binary analysis. Symbolic execution finds hidden paths. github.com/angr/angr Drop this into your Claude, then watch Opus 5.5 take apart any app you own below.
codila
@0xCodilaOct 12
REA is most dangerous thing on the internet right now... It lets your AI agent reverse engineer any app, down to the binary If you use it correctly, you can rebuild any feature you pay for, for FREE In this article, I show you how:

Reverse Engineering: How to Get Anything for FREE With Most Dangerous Repo & Opus 5.5 (Full Course)

Article · @0xCodila
17
102
3
594
36,230
🔥🚒 pfSense Firewall 0day - 5 RCEs! 🚒 🔥 Exploits - github.com/abraxas/pfsense-q… - github.com/abraxas/pfsense-o… - github.com/abraxas/pfsense-c… - github.com/abraxas/pfsense-l… - github.com/abraxas/pfsense-i… Writeups - abraxaslabs.tech/research/pf… - abraxaslabs.tech/research/pf… - abraxaslabs.tech/research/pf… - abraxaslabs.tech/research/pf… - abraxaslabs.tech/research/pf…
🤖 Made with AI
15
200
16
804
73,256
quincunx33/ai-jailbreak: A collection of jailbreak prompts and exploit techniques for local and frontier AI models, with modern methods for Qwen3.5, Gemma 4, Llama 4, Kimi K3, GPT-OSS, GPT-5.x, Gemini 3.x and Grok 4.x. For red-... github.com/Quincunx33/Ai-jai…

GitHub - Quincunx33/Ai-jailbreak: A collection of jailbreak prompts and exploit techniques for...

A collection of jailbreak prompts and exploit techniques for local and frontier AI models, with modern methods for Qwen3.5, Gemma 4, Llama 4, Kimi K3, GPT-OSS, GPT-5.x, Gemini 3.x and Grok 4.x. For...

github.com
3
189
4
1,104
57,226
New Uncensord Cyber models (4B–20B) is a family of cybersecurity-expert small language models you can run locally. A 20B specialist trained for threat management and SecOps. - CTI questions and ATT&CK mapping - CVE / ticket to CWE - detection and mitigation suggestions - investigation support - Trained with SecKnowledge 2.0 expert-in-the-loop reasoning formats plus document/web grounding. - On correlating vulnerabilities and bug tickets with weaknesses, - the 20B model outperforms GPT-4o, o1, o3-mini, and Sec-Gemini, ranking first. - multi-step grounding, and higher-fidelity reasoning traces for security tasks. Treat outputs as advisory. Human review still required. - huggingface.co/cyber-pal-sec…
Md Ismail Šojal 🕷️
@0x0SojalSecSep 29
The Cyber-OSINT model that You can run locally. - MoE (26B total, 4B active, 262K ctx) trained on 6,500 OSINT/CTI instructions. - SFT for cyber threat intel and investigative work. - Threat-actor attribution. - IoC pivoting. Geolocation. - Admiralty source grading. - 262K context.
1
24
169
9,522
swem retweeted
CR1337
@CR1337Sep 30
Always one step ahead, anon
32
91
11
1,276
41,525
Prompt injection cybershujin.github.io/Prompt…
2
27
149
6,912
swem retweeted
lumxss
@bkdgiffugSep 29
港大开源的 DeepTutor，把 AI 家教做得有点像“私人导师”。 它不是简单的一问一答，而是先把你的教材、论文、笔记喂进去，再根据这些资料陪你学习。 ① 深度解题，把复杂推理拆成步骤 ② 自动出题、批改，针对薄弱点生成练习 ③ 自动调研复杂主题，整理成结构化报告 ④ 把抽象知识变成图表、动画和互动内容 ⑤ 记录学习进度，动态调整学习路径 支持 RAG、本地模型、Skills 和长期记忆，也能 Docker 自托管，Apache 2.0 开源。 GitHub 已经 3.6 万+ Stars，想搭一套属于自己的 AI 学习系统，可以看看。 🔗 github.com/HKUDS/DeepTutor
8
33
1
108
7,662
Top 10 OpenCode Skill Repos: 1. Superpowers - github.com/obra/superpowers 2. Ponytail - github.com/DietrichGebert/po… 3. UI/UX Pro Max - github.com/nextlevelbuilder/… 4. Graphify - github.com/Graphify-Labs/gra… 5. Caveman - github.com/JuliusBrussee/cav… 6. Addy Osmani Skills - github.com/addyosmani/agent-… 7. Understand Anything - github.com/Egonex-AI/Underst… 8. Awesome Claude Skills - github.com/ComposioHQ/awesom… 9. Archify - github.com/tt-a1i/archify 10. Impeccable -github.com/pbakaus/impeccabl…
79
627
19
4,701
266,604
Fully Uncensored Cyber-Ornith-1.5-9B-OBLITERATED You can run in your laptops. - A 9B cyber agent that actually talks to your multi-turn terminal ops - Trained on 45k cybersecurity trajectories data - Zero refusal - 100% CyberSecEval triage - 95% Hermes tool-call AST. - +20% GPQA Diamond CoT - Keeps 'thinking' before it acts vuln triage. - Local threat-hunting sidekick. - Agentic cybersecurity model. - CLI tasks, function-calling,
13
73
2
580
56,048
swem retweeted
lumxss
@bkdgiffugSep 25
AI Agent 跑长任务最烦的，就是聊着聊着“失忆”，每次都得重新塞历史上下文，Token 也跟着暴涨。 SimpleMem 做的就是一套开源长记忆系统。 它不会把整段对话原样存下来，而是先整理成独立事实，需要时再通过语义检索找回来。 官方测试里，Token 消耗最高可减少约 30 倍，LoCoMo F1 提升 26.4 个百分点；现在还支持图片、音频、视频等多模态记忆。 另外还有自进化检索机制，会根据失败案例调整策略。 不过它并非完全本地运行，内部 LLM 调用仍需要 OpenAI 兼容 API Key。 🔗 github.com/aiming-lab/Simple…
18
11
47
3,554
WTF！太激动了，手现在都在颤抖 这 7 个 GitHub 项目让我兴奋一天，太上隐了！ 全是tm搞钱工具，找资源、找 API、抓数据、跑模型、搭 Agent、做 AI 产品，基本覆盖完整链路： 1、Awesome（50.9 万+ Stars）｜AI 资源地图 整理所有了AI、开发框架、自动化工具、学习资料等海量资源，相当于一个 GitHub AI 导航站，搜索效率提升 10 倍 github.com/sindresorhus/awes… 2、Public APIs（48.2 万+ Stars）｜AI 产品数据宝库 天气、金融、地图、图片、游戏大量公开 API，让你不用花几天时间到处找数据源，开发周期直接缩短 50%+ 做 AI 应用都不用去爬数据了 github.com/public-apis/publi… 3、Scrapling（8.3 万+ Stars）｜让 AI 获取互联网信息 很多 Agent 最大的问题不是不会思考，而是没有实时数据。 Scrapling 可以帮 AI 快速获取网页内容，把新闻、资料、竞品信息接入工作流，信息收集效率提升 5 倍+。 github.com/D4Vinci/Scrapling 4、Free for Dev（13.8 万+ Stars）｜开发者省钱神器 服务器、数据库、云服务一直是 AI 创业成本 这个项目整理大量免费开发资源和优惠方案，个人开发者可以少花大量试错成本，节省 100 + 美元级投入 github.com/ripienaar/free-fo… 5、Ollama（18.1 万+ Stars）｜本地 AI 基础设施 直接把开源模型跑到自己的电脑上，搭私人助手、知识库、Agent 工作流，可以减少 90% API 成本 github.com/ollama/ollama 6、Open WebUI（15.2 万+ Stars）｜免费打造自己的 ChatGPT 直接拥有网页端 AI 工作台，聊天、文件分析、多模型切换、知识库全部整合，让本地 AI 从“技术玩具”变成真正生产工具 github.com/open-webui/open-w… 7、Dify（15.7 万+ Stars）｜AI 应用开发加速器 之前需要几周验证的 AI 产品想法，现在几个小时就能做出 MVP，开发效率直接提升 10 倍 不会复杂代码，也能快速搭建 AI Agent、知识库和自动化工作流。 github.com/langgenius/dify
govin.eth | G哥
@goan999999Sep 24
卧槽，兄弟们！Muse 这波已经被 GitHub 玩明白了 Muse爆火！我找到 3 个专门给 Muse.ai 用的免费开源项目，直接把 Muse 从“聊天工具”升级成更强的 Agent 工作流 1、让 Muse 真正帮你干活：任务模板库 里面整理了 100+ 个 Agent 工作流模板，覆盖研究、自动化、信息整理等场景，你想要的都有！ 不用每次重新设计 Prompt，直接复制修改，几分钟就能搭建自己的 AI 助手 github.com/Anil-matcha/aweso… 2、让 Muse 进入终端：Muse CLI 支持聊天、目标管理、动态查看、Ideas 和 Sessions 功能，让开发者日常调用效率提升 50%+ 把 Muse 直接搬到 Terminal，不用频繁切换网页了 github.com/nikships/muse-cli 3、让其他 AI 调用 Muse：Muse MCP 通过 MCP 协议连接 Muse，让 Codex、Claude Code 等多个 Agent 可以调用 Muse 能力 把单个 AI 助手升级成多 Agent 协作流程，扩展效率提升 10 倍 github.com/kleprevost/muse-m…
81
516
13
2,431
272,847
Here is list of 10 FREE hosting providers, to Host your Static Websites - pages.cloudflare.com - pages.github.com - docs.gitlab.com/user/project… - netlify.com - vercel.com - firebase.google.com/products… - render.com - surge.sh - codeberg.page - neocities.org
17
79
2
444
16,731
Vibe-Hardware is not as simple as it looks. It’s not the flashy demo where AI generates a nice-looking PCB in 20 seconds and calls it done. The hard part starts after that. 1) The circuit has to actually work 2) The PCB has to be valid 3) The enclosure has to fit and the parts have to exist 4) And then there are about 50 other tiny details that somehow turn into VERY expensive problems That’s why we’re approaching vibe hardware differently with heypcb.ai. We’re building an AI-native EDA + MCAD platform, where AI works inside the actual design environment and understands both the electronics + mechanical design as you iterate. Generating hardware is cool. Building hardware you’d actually trust enough to manufacture is the real problem.
26
55
5
553
26,377
Load more