Stripe Says Kai Runs on 1,000 Skills. Its Partner Says Quality Drops Past 150
Stripe says its internal agent is connected to 1,000+ skills. Its infrastructure vendor reports quality degrading past 150. A close read of both posts.
Stripe says its internal agent is connected to 1,000+ skills. Its infrastructure vendor reports quality degrading past 150. A close read of both posts.
A third-party benchmark finds one prompt sentence cuts invented fields from 70.7% to 20.2% across 16 models, and that a paid API on a free tier fabricated more than 13 of them.
Meta’s engineering write-up for Muse states its premise in one line: two isolated security domains on one box, not an LLM powered agent with root. A walk through what that means in practice.
Linear’s engineers found that faster code generation just moved the wait. An account of what happened when the bottleneck shifted downstream.
An open-source challenger to Jev ships with a comparison table. The accuracy claim does not survive reading the rest of the page, but the structural argument does.
A published critique argues that calling OpenAI’s agents rogue lets the company off the hook. Checked against the two primary incident reports, its language claim holds and its central factual claim does not.
Claude Code’s AGENTS.md loader was gated behind a remote feature flag, so telemetry-off sessions silently skipped the file. Anthropic fixed it in about five hours and the changelog does not say so.
OpenAI’s GPT-6 Sol and Luna announcement is a cost-efficiency pitch. Its own footnotes concede that the closest competitor’s cost is the one figure not reported.
A week after launch is when the real question gets answered: does anyone build on it. For Jev the early answer is code review, model routing and linting - places where which one matters more than write me something.
Every model you have used generates text one token at a time, even when all you needed was yes or no. TypeSafe AI built one that returns typed values instead.