AI Builders

@theaibuilders

A community where builders turn ideas into real products. JP ACCT @aibuildersjapan linktr.ee/theaibuilders

theaibuilders.dev

Joined September 2025

772 Following

1,540 Followers

Replies Reposts Media

Pinned
AI Builders
@theaibuilders
Oct 31
Cursor: a tampered Claude Code extension (2.1.312/2.1.313) reached up to 5,000 users and appended an SSH key to authorized_keys on load. The extension marketplace is the trust boundary. "Official" is a label, not a control.
Cursor
@cursor_ai
Oct 30
We identified a security incident affecting a modified Claude Code VS Code extension accessed through our Cursor IDE extension marketplace, impacting at most 5,000 users. This malicious version of the extension was not developed by Anthropic. It does not impact other extensions
1
AI Builders
@theaibuilders
Nov 6
GPT-6 Cyber found 81 unknown Firefox flaws and working exploits for 948 Linux kernel bugs. Mozilla is patching. The model is limited to vetted security customers. We built a machine that outruns the patch pipeline, then handed it to very few people. Safety, or triage?
1
AI Builders
@theaibuilders
Nov 5
OpenAI spent ~10 weeks building Pando, a Git backend, because Codex agents strained push latency. Anthropic swapped Git for Scion, claiming 1/200 server cost. AI coding's bottleneck isn't the model. It's the 20-year-old plumbing agents write through. What gets rewritten next?
AI Builders
@theaibuilders
Nov 4
56% of US voters said data centers were a major factor in their midterm vote. Texas went blue on a moratorium platform. The AI buildout is now a land-use fight, and those are won by whoever shows up at the zoning meeting. Your compute roadmap has a precinct map now.
AI Builders
@theaibuilders
Nov 3
Epoch AI gave open-weight GLM 5.3 a Firefox build and the public patch feed. Working exploits for 6 straight releases, ~$1,000 each, zero refusals. Chrome held. Everyone blames open weights. The real bug is shipping the fix before the fix ships.
Rohan Paul
@rohanpaul_ai
Nov 2
In Epoch AI tests, Z .ai’s open-weight GLM 5.3 turned unshipped Firefox security fixes into working attacks (i.e. built working exploits) on 6 consecutive releases. Those fixes sit in public code before reaching the stable browser, disguised as routine changes to slow attackers.

AI Builders

AI Builders’s posts