8 min read
Claude Opus 5 Benchmarks: What the Numbers Actually Show
Opus 5 posts 79.2 on SWE-bench Pro against Opus 4.8 at 69.2. What the launch ratios mean, and…
The Lab · Written in the open
Working notes from the studio. The tools I build, the workflows I keep, the mistakes I paid for.
469 entries · latest 2026-10-22
8 min read
Opus 5 posts 79.2 on SWE-bench Pro against Opus 4.8 at 69.2. What the launch ratios mean, and…
9 min read
Anthropic shipped Claude Opus 5 on July 24, a step-change model with a 1M context window at half…
10 min read
Why I build five focused tools instead of one sprawling product, and what Git Dojo, OhNine, and Statusline…
9 min read
Anthropic updated Claude's voice mode with Opus and Sonnet, mid-call model switching, and real actions in Gmail, Calendar,…
9 min read
Anthropic shipped Claude Security in beta, a multi-agent scanner that turns vulnerability findings into reviewed patches inside Claude…
9 min read
Claude Blueprint packages a full Claude Code setup, a CLAUDE.md template, 6 commands, 6 skills, and 3 hooks,…
10 min read
OhNine is a free menu bar app that tracks Claude session and weekly usage limits and warns you…
9 min read
Why one caption pasted everywhere underperforms, and what building a platform-aware caption tool taught me.
9 min read
Statusline Builder configures a Claude Code statusline with 33 elements, free forever. Here is why I built it…
10 min read
Git Dojo teaches git through real terminal commands, not menus. Why I built it that way and what…
9 min read
Kimi K3 topped a coding arena, Gemini 3.5 Pro slipped, and OpenAI shipped GPT-Live. A builder's read on…
9 min read
Claude Code's new screen reader mode swaps boxes and spinners for plain, linear text. Why accessibility in a…