Secure, Scalable & Compliant DevSecOps, CyberSecurity, AI tooling, SRE & FinOps Solutions. Connect via https://teapaw.org/t.co/CaUilu5rdL or email for a consultation.
Bengaluru, India
Joined May 2015
- Tweets620
- Following422
- Followers43
- Likes3.1K
ProDevOps retweeted
This is the state of security in India, largest broker @zerodha with AUM of around $60B offers whopping $1.5k for a P1 bug report, can you imagine?
Just now CBSE, largest and government owned education department leaked all the data and when @ni5arga reported they did not bother.
Sometime ago (2 years) I found a bug in some government medical website with which EVERY SINGLE INDIAN who ever used government services, their blood report can be accessed with just mobile number, I went out of my way to ping all the associated people on linkedin, emails and everything, but 2 years later it's still live.
Now people + gov will try to hunt down/ put cases on this kid for reporting issues publicly.
I had hacked CBSE's OSM (On-Screen Marking Portal) in February and had reported the vulnerabilities to CERT-In, but they were unable to patch most of them.
I've written a detailed blog post about it here: ni5arga.com/blog/posts/hacki…
ProDevOps retweeted
Do you guys want to know how the biometric system is bypassed/ hacked? It's ingenious. I know it because advised on one such case.
How do you stop exam paper leaks? A tech- based reform required.
Under the leadership of DGP,Haryana, Sh. Ajay Singhal — an innovator and reformer in policing — Haryana Police rebuilt the B1 promotion exam from scratch. Zero paper. Zero human contact. Zero leaks.
Between 2018-2024: 7 exams | 34,278 constables appeared | 14,000 unique question papers | Each candidate got a different set of 140 Qs | Instant email results | CCTV + biometric + firewalls.
Result: Zero cheating. Zero litigation.
First police force in India to do this.
“Every leaked paper is a letter of rejection to fairness.” - DGP Ajay Singhal
When will, planning & technology combine, trust is restored. This is how you dismantle the exam mafia.
#TechForTrust #HaryanaPolice #ExamReforms #DGPInitiative
#PoliceReforms
#Fairness
#TribuneCHD
@cmohry @NayabSainiBJP @csharyana @DgpChdPolice @dgpup @DGPPunjabPolice @dgpapofficial @DGPMaharashtra @CPDelhi @thetribunechd @htTweets
ProDevOps retweeted
A new generation of people are realising how messed up CERT-IN and India's cybersecurity is.
ProDevOps retweeted
I once criticized CERT-In on LinkedIn and got calls from terrified employers (past & then present) asking me to remove it.
I said no and proceeded to change all my employment history to ‘Confidential’ to ease their worries.
The reason they panic is because they don’t want to lose their “CERT-In Empanelment”.
CERT-In Empanelment is one of the biggest scams in the Indian cyber security industry.
CERT-In makes you go through several stages of tests (all worthless btw) and then “certifies” you as an empaneled auditing firm.
This status then allows you to bid on government contracts for cyber security projects along with enabling you to serve compliance customers under regulatory bodies.
If, for whatever reason, CERT-In decides to revoke this empanelment, the firm would lose majority of its business.
That’s how CERT-In keeps all the major cyber security firms in India under their thumb.
ProDevOps retweeted
You can now run a full Linux operating system inside a 6mb PDF.
Someone embedded a RISC-V emulator inside a standard document. You don't need a virtual machine, just a PDF reader.
→ Runs interactively inside the file.
→ Powered by a tiny RISC-V emulator.
→ The entire OS fits in just 6MB.
ProDevOps retweeted
Mercor AI has allegedly been breached by Lapsus
939GB of source code
4TB of data in total
All data from their TailScale VPN
@mercor
ProDevOps retweeted
SOMEONE DROPPED A FULLY JAILBROKEN, OPEN-SOURCE, PRE-BUILT IOS 26 VIRTUAL MACHINE
Repo: github.com/34306/vphone-aio
Readers added context they thought people might want to know
The "someone" who released this is @Little_34306 (as seen on the linked GitHub: github.com/34306). The OP regularly fails to credit the work they post.
ProDevOps retweeted
PSA: If you've been running out of Claude session quotas on Max tier, you're not alone. Read this.
Some insane Redditor reverse engineered the Claude binaries with MITM to find 2 bugs that could have caused cache-invalidation. Tokens that aren't cached are 10x-20x more expensive and are killing your quota.
If you're using your API keys with Claude this is even worse. This is also likely why this isn't uniform, while over 500 folks replied to me and said "me too", many (including me) didn't see this issue.
There are 2 issues that are compounded here (per Redditor, I haven't independently confirmed this) :
1s bug he found is a string replacement bug in bun that invalidates cache. Apparently this has to do with the custom @bunjavascript binary that ships with standalone Claude CLI.
The workaround there is to use Claude with `npx @anthropic-ai/claude-code`
2nd bug is worse, he claims that --resume always breaks cache. And there doesn't seem to be a workaround there, except pinning to a very old version (that will miss on tons of features)
This bug is also documented on Github and confirmed by other folks.
I won't entertain the conspiracy theories there that Anthropic "chooses" to ignore these bugs because it gets them more $$$, they are actively benefiting from everyone hitting as much cached tokens as possible, so this is absolutely a great find and it does align with my thoughts earlier.
The very sudden spike in reporting for this, the non-uniform nature (some folks are completely fine, some folks are hitting quotas after saying "hey") definitely points to a bug.
cc @trq212 @bcherny @_catwu for visibility in case this helps all of us.
ProDevOps retweeted
Ironic how Anthropic sells Claude Code security reviews positioned as something v powerful (costing $15-25 per PR review), and being clear they use it on all PRs... then leaking all of Claude Code's code thanks to publishing their sourcemap.
AI won't save you from yourself!
ProDevOps retweeted
🚨 ZERODAY: ImageMagick 🚨
Our autonomous pentester pwn.ai just dropped multiple zeroday chains in ImageMagick that achieve RCE and File Leak from a single .jpg or .pdf file, bypassing EVERY security policy (Default, Limited, AND Secure). 🤯
💥 Affects Ubuntu, Debian, WordPress & millions of servers globally. Happy Monday and Happy Hunting! 🥰
pwn.ai/blog/imagemagick-from…
ProDevOps retweeted
Holy shit… 4TB of job INTERVIEWS, RESUMES, and AI SECRETS just leaked from a $10 BILLION startup.
MERCOR uses AI to match people with high-paying gigs and literally records your face, voice, KYC docs and keeps it forever to “match” you with jobs.
Your FACE + VOICE + ID could be on the dark web right now.
Hackers (Lapsus$) claim they stole 4 TB through the company’s Tailscale VPN:
- 211 GB of candidate resumes and personal data
- Terabytes of video interviews + passport + KYC docs
- Almost 1 TB of source code
damn that Claude Mythos "God Tier" at Cyber stuff got into the wrong hands lol
Mercor AI has allegedly been breached by Lapsus
939GB of source code
4TB of data in total
All data from their TailScale VPN
@mercor
ProDevOps retweeted
We asked Claude to find a bug in Vim. It found an RCE. Just open a file, and you’re owned. We joked: fine, we’ll switch to Emacs. Then Claude found an RCE there too.
Full story: blog.calif.io/p/mad-bugs-vim…
ProDevOps retweeted
New supply chain attack this time for npm axios, the most popular HTTP client library with 300M weekly downloads.
Scanning my system I found a use imported from googleworkspace/cli from a few days ago when I was experimenting with gmail/gcal cli. The installed version (luckily) resolved to an unaffected 1.13.5, but the project dependency is not pinned, meaning that if I did this earlier today the code would have resolved to latest and I'd be pwned.
It's possible to personally defend against these to some extent with local settings e.g. release-age constraints, or containers or etc, but I think ultimately the defaults of package management projects (pip, npm etc) have to change so that a single infection (usually luckily fairly temporary in nature due to security scanning) does not spread through users at random and at scale via unpinned dependencies.
More comprehensive article:
stepsecurity.io/blog/axios-c…
🚨 CRITICAL: Active supply chain attack on axios -- one of npm's most depended-on packages.
The latest [email protected] now pulls in [email protected], a package that did not exist before today. This is a live compromise.
This is textbook supply chain installer malware. axios has 100M+ weekly downloads. Every npm install pulling the latest version is potentially compromised right now.
Socket AI analysis confirms this is malware. plain-crypto-js is an obfuscated dropper/loader that:
• Deobfuscates embedded payloads and operational strings at runtime
• Dynamically loads fs, os, and execSync to evade static analysis
• Executes decoded shell commands
• Stages and copies payload files into OS temp and Windows ProgramData directories
• Deletes and renames artifacts post-execution to destroy forensic evidence
If you use axios, pin your version immediately and audit your lockfiles. Do not upgrade.
ProDevOps retweeted
If this is not a meme it’s peak comedy.
Anthropic didn’t get hacked.
They just shipped their full 57 MB Claude Code CLI source, every prompt, every system instruction, every secret, to npm.
The “safety-first” $10 B lab forgot a oneline .npmignore.
Proof that AI labs are now so mediocre at software engineering, the chef is literally serving the entire kitchen.
When your chef is serving tables… the restaurant is cooked. 💸🔥
Claude Code 泄露了全部源码——不是被黑客攻破，是 Anthropic 自己把 source map 打包进了 npm 发布物。
一个 57MB 的 cli.js.map 文件，里面藏着 4756 个源文件的完整内容。其中 1906 个是 Claude Code 自身的 TypeScript/TSX 源码，剩下 2850 个是 node_modules 依赖。
提取方法极其简单：cli.js.map 本质就是一个 JSON，里面有两个关键数组——sources（文件路径）和 sourcesContent（对应的完整源码）。两者索引一一对应。不需要反编译，不需要反混淆，sourcesContent 里存的就是一字不差的原始代码。提取脚本见文末。
从还原的源码可以看到：Claude Code 用 React + Ink 构建 CLI 界面，核心是一个 REPL 循环，支持自然语言输入和 slash 命令，底层通过工具系统与 LLM API 交互。架构设计、系统提示词、工具调用逻辑，全部一览无余。
这件事的本质是一个经典的安全疏忽：source map 是开发调试用的，包含从变量名到注释的所有信息，不应该出现在生产发布物中。Anthropic 后来意识到了这个问题，移除了 source map，GitHub 上提取源码的仓库也被 DMCA 了。但早期版本的 npm 包已经被存档，源码早就在社区流传。
给所有发布 npm 包的开发者提个醒：发布前检查你的 .map 文件。一行 sourcesContent 就能让你的所有代码公之于众。
gist.github.com/sorrycc/ec29…
ProDevOps retweeted
Conspiracy theory: They leaked it so that real humans can finally review the code. Since they do not read or review the code themselves at all. Now they ask us to do it for them for free.
Claude code source code has been leaked via a map file in their npm registry!
Code: pub-aea8527898604c1bbb12468b…
ProDevOps retweeted
I created documentation over Claude Code's Codebase, which explains
- Its pipeline
- How it works
- How it handles Context
- How it handles Memory
& More
Read it here - mintlify.com/VineeTagarwaL-c…
Claude code source code has been leaked via a map file in their npm registry!
Code: pub-aea8527898604c1bbb12468b…
ProDevOps retweeted
Claude code source code has been leaked via a map file in their npm registry!
Code: pub-aea8527898604c1bbb12468b…
ProDevOps retweeted
Superterm is now free for personal use.
Manage your agents from one place. Unblock them from your phone.
Let's ship. Link below 👇
ProDevOps retweeted
everyone thinks that entire industries can be destroyed by a markdown file. it's true but doesn't give the whole picture.
agents are coming face to face with a million edge cases that humans deal with without even noticing them. all these edge cases have to first be discovered and then put in that markdown file. without this, the outcomes from the markdown file will not be worth paying for.
the people and teams that codify their knowledge in precise enough detail that an agent can deliver comparable outcomes are going to have valuable markdown files. and you will have to pay to access these skills.
these skills will only be delivered via an agentic api where you ask the questions and get the results without being able to see the working of the model. yes even tool calls and results will be hidden away.
it will start to look like the web all over again - dumb clients doing user facing tasks and highly intelligent servers doing the real cognitive heavy lifting.
