I ship things people love using • mts @mem0ai - https://xitter.cf/t.co/n9C3OZSXkr (30k+ mau) - https://xitter.cf/t.co/oPkkjxsyZC (20k+ downloads) - https://xitter.cf/t.co/acE2EdY8ZX
Joined March 2023
- Tweets17.7K
- Following971
- Followers10.3K
- Likes117K
Kartik retweeted
cricket 07 with an online multiplayer mode. play with your friends
all you need is a web browser
link: cricket07.dhaiwat.com
This is a bad month to be a software engineer.
I feel after this we will have good guardrails for orchestration engines.
Kartik retweeted
October has been an absolute nightmare for software
> Railway - customer secrets exposure
> LiteLLM - malicious PyPI + Docker releases
> Zeabur - infrastructure breached
> OpenCode - AI API keys exposed
> OVHcloud CLI - official binaries backdoored
> Rank Math PRO - malware in official plugin
> TeamUpdraft - customer data compromised
> ConnectWise - cloud backup identities compromised
> Tensorlake - npm package infected
> SubQuery - backdoored npm package
> GhostAction - GitHub repos with malicious workflows
> PHP - critical RCE vulnerabilities
> Node.js - security vulnerabilities patched
> eMudhra - certificate authority keys compromised
> Tailscale - eMudhra incident
At this point, the biggest security risk might not be the code we write, but the code we trust...
This was the final nail in the coffin.
Anyways I never liked Railway in the first place.
Had really bad experience with them.
An urgent notice from the Railway team
blog.railway.com/p/security-…
I don't know why people were using railway in the first place.
There infra is not reliable and they do not take any responsibility for it as well.
deepseek open-sourced the inference engine behind its production services.
DeepInfer includes deterministic sampling, structured outputs and async scheduling. the current release supports offline inference for V4.1-Flash.
i'd read this for the serving decisions. downloading the weights was only the first step.
Openclaw
Hermes
Grok bot
Instinct
Dots
And now duet
We are in the era of personal agents.
we gave ai the ability to write entire apps and then made it ask us to check the verification email lmao
agentmail giving agents their own inbox makes sense.
if mine can handle signups and chase invoices, i'm so here for it.
🤝 Paid partnership
Today, @agentmail has processed over 1 Billion emails.
A year ago, we processed under 20 million.
So we figured it was time for AgentMail to grow up a little too :)
I think more companies from startups to MNCs are going to move toward fewer third-party dependencies in their projects with supply chain attacks happening so often this could become the norm pretty soon.
Kartik retweeted
Today, @agentmail has processed over 1 Billion emails.
A year ago, we processed under 20 million.
So we figured it was time for AgentMail to grow up a little too :)
an autonomous agent that waits forever is not autonomous. it is just stuck quietly.
claude code 2.1.301 added a retry watchdog for max-usage-limit waits. it also sends classifier verdict, mode and category through telemetry, so teams can see why a safety path triggered instead of treating it as a mystery.
the release includes more safety fixes too.
these are not flashy model capabilities, but they decide whether a long coding run can be trusted unattended. give every wait a limit, expose every policy decision, and make the failure visible enough for a human to act.
"Sir openai models helped in discovering the PHP vulnerabilities. We are cooked "
OpenAI models helped researchers—operating as part of our Patch the Planet initiative—discover critical security flaws in PHP, an open-source programming language used to run an estimated 70% of websites.
Over the past two months, we led a coalition of maintainers, researchers, and security providers to prepare fixes and protections ahead of disclosure. This duration was necessary because the fix was quite intricate and required a major change to the PHP engine. Every site using PHP can now safely upgrade their version to stay secure, in addition to applying mitigations from various security vendors.
Due to the severity of these flaws and PHP’s widespread use, we consider this bug to be one of the most impactful vulnerabilities in recent memory. It also marks a new phase, where AI can find security bugs with internet-scale impact.
This is an example of the Defender’s Window in action: to our knowledge, frontier AI was able to find this vulnerability ahead of attackers, giving defenders time to prepare.
openai.com/index/coordinated…
Readers added context they thought people might want to know
OpenAI may claim a role in the coordinated response and patching process. But the initial discovery and exploit development, however, occurred several months earlier by an independent researcher.
x.com/lcfr_hax/statu…
Kartik retweeted
Today, I'm launching ihateDesmos.com:
- SO Powerful than any other Traditional Mathematics tool out there
- Draw any shape by hand, and spinning circles redraw it.
- Zoom into the Mandelbrot set billions of times and it stays sharp.
- See step-by-step solutions, each checked before it's shown to you.
- Factor, solve and simplify expressions exactly, not just as decimals.
- Find derivatives, integrals, limits and Taylor series.
- Write notebooks that mix text, maths, live graphs and Python code.
- Leave comments and roll back to any earlier version of your work.
- Get answers to 50 digits when you need extreme precision.
- Calculate with units: 9.81 m/s² × 3 s gives 29.43 m/s.
- Export animations as GIF or video, plus PNG, SVG, PDF and LaTeX.
- Bring your graphs over from Desmos and GeoGebra in one step.
- Tricky curves are checked pixel by pixel, so nothing is missed or faked.