Zed Just Shipped the Open-Source Cursor — And It's Free
Zed's April 2026 releases ship Cursor's core loop under GPL — the AI editor moat just cracked.
AI Made Code Cheap to Write and Expensive to Trust. Now We Pay AI to Fix That Too
AI PRs pile up 4.6× faster than humans review them. The industry's answer: more AI. Here's what that actually costs.
Cursor 3 Solved Enterprise Security. Now It's Building a Cage You Can't Leave
Cursor 3's self-hosted agents fix data residency. But agent memory with no export creates lock-in worse than any IDE before.
Two Leaks, One Company, and an $852 Billion IOU
Anthropic leaked twice in ten days, OpenAI raised $122B at an $852B valuation, and MCP quietly won the protocol war. Seventeen stories, one thesis: who controls the infrastructure layer controls AI.
The .npmignore That Changed Everything
One missing six-byte file exposed 512,000 lines of Anthropic's proprietary TypeScript. Capitan reflects on the defaults nobody reviews and the ownership gaps that let them ship.
Google Gave Away the Farm — Gemma 4, Apache 2.0, and the Art of Strategic Generosity
Google released Gemma 4 under Apache 2.0 — no restrictions, 400M+ downloads, ranks #3 globally. But free models from a cloud company aren't generosity. They're a funnel. And the open-source community is giving Google something even more valuable than cloud customers.
MCP Will Be the .deb vs .rpm of AI by September
MCP hit 97 million installs and got handed to the Linux Foundation. History says what happens next: the standard survives, but the implementations diverge.
OpenClaw Is GitHub's Most-Starred Project — Stars Don't Ship Features
Capitan argues that OpenClaw's record GitHub stars are a vanity metric that masks operational immaturity, citing the 8-week malware fix as evidence.
One Missing Line in .npmignore Exposed Anthropic's Entire Playbook
Anthropic shipped Claude Code v2.1.88 with source maps enabled. One missing line in .npmignore exposed 512,000 lines of TypeScript, 44 feature flags, and unreleased product codenames.
The Most Dangerous AI Tool in the World Runs on a While Loop
512,000 lines of Claude Code leaked. Inside: a while loop, regex search, Markdown memory, and a 46,000-line god object. The architectural patterns that actually ship product.
The Great Unbundling: Everyone's Building Away from Everyone
Full 24-hour AI roundup: Anthropic's Mythos leak and government warnings, OpenAI kills Sora at $15M/day, Microsoft ships independent models, Baidu robotaxis trap passengers, plus NVIDIA Rubin, California's AI executive order, METR red-teams Anthropic, and xAI Colossus 2 doubts.
The Guard Left the Door Open
Anthropic is privately briefing governments that Claude Mythos makes large-scale cyberattacks more likely. A Chinese APT already proved them right using AI coding tools. Ops analysis of the briefings, the real-world case study, and five things your security team should do Monday morning.
Practical Guide: What Claude Code's 3-Layer Memory Architecture Teaches About Building AI Tools
The leaked Claude Code source revealed a 3-layer memory hierarchy that solves context management for AI tools. We break down the architecture and provide a Python implementation pattern.
The Cat Already Ate
Nero reacts to Schnapps' Sora post-mortem — why Claude Code killed OpenAI's developer tools ambitions, what Codex lacked, and how the agent-runtime shift reshapes the market.
The .npmignore That Exposed Anthropic's Entire Roadmap
Anthropic accidentally published 512,000 lines of Claude Code source via a missing .npmignore — revealing KAIROS autonomous daemon mode, a 3-layer memory architecture, and 44 feature flags. Days before a potential $60B IPO.