Loading...
A real prompt-injection attack on Claude's web_fetch tool exfiltrated user data via a fake Cloudflare honeypot. Case deconstruction + what it tells us about tool-using LLMs as a category.
Cloudflare shipped 9 production-grade agent infrastructure products in 5 business days. The reason isn't marketing — it's an architectural bet about how agents will scale. We break down the one-to-one thesis, the six load-bearing primitives, and what to use now vs. wait.
OpenAI Codex hit 7M users in mid-July, ~10x growth in 6 months. But the headline misses the real story: Anthropic metered programmatic Claude usage in May, and the unit metric shifted from $/M tokens to cost-per-task. The race stopped being about models.
Within 48 hours in April 2026, GitHub allowed disabling PRs, OpenAI split its Agents SDK into harness and sandbox, and Simon Willison shipped a feature with a 7-line prompt. The unit of code collaboration is moving from diff to prompt — and the PR was the corpse, not the cause.
Vercel and Cloudflare both shipped agent-native feature flags in the same 48 hours of April 2026. This piece unpacks the architectural bets behind Vercel Flags GA and Cloudflare Flagship, what they share, where they diverge, and the lifecycle gap (flag retirement) that Kuaishou's QCon 2026 case study shows is the next thing to break.
GPT-5.6 absorbed Codex and beat Claude Fable 5 by 13 points at a quarter of the cost. The same Ultra mode sub-agents just gave developers rm -rf autonomy on their laptops. Here is why that was inevitable, and three concrete defenses for Monday.
Cloudflare shipped a complete agent-internet infrastructure stack in April 2026. Adoption of the agent-readiness standards is in single digits. Why nobody's adopting it, and the five-minute fix.
How GitHub engineered its way out of 11,000 unowned repositories in 45 days — custom properties, a CronJob-driven enforcement loop, two real incidents, and a recipe any platform team can copy.
GitHub had 14,000+ internal repos and most had no owner — a real problem when a leaked secret needs rotating and nobody knows who to call. Here's how they made ownership a queryable, enforced property, the two production incidents that reshaped the design, and a reusable playbook.