Claude Code's limits just doubled


Claude Code's limits just doubled
Plus the silent 1-hour cache bug in 2.1.131, Episode 2 of the production series, and every WotAI live session now on YouTube.  ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌
WotAI

Claude Code's limits just doubled

Hey Reader,

Eight days, four posts. Big week for Claude Code: 5-hour rate limits doubled (and Anthropic signed for an entire SpaceX data center to back it up), 2.1.131 quietly fixed a 1-hour prompt-cache bug that had been silently overcharging anyone running long-TTL cache control, Episode 2 of the production series shipped, and the full WotAI live-session archive is now public on YouTube.

Four reads from the last 8 days. I'd go in this order.

Episode 2 – Claude Code's four execution modes shipped a 3-week handoff in 18 minutes

Production Claude Code Episode 2 cover - SHIFT+TAB and the four execution modes

This is Episode 2 of the production series. Most Claude Code users press shift+tab once, see "accept-edits," and never come back. There are four modes total: default (y/n on every action), accept-edits (bulk apply, no per-file prompts), plan (researches and proposes – the only mode that physically can't write to your files), and auto (unattended within a /permissions allowlist you set).

The leverage isn't any single mode. It's the sequence: plan first to write the rollout doc without granting permissions, then accept-edits to scaffold, then /permissions allowlist, then auto mode to run the migration and wire up integrations. I had a Claude Design handoff that had been sitting in my repo for three weeks. Used the four modes in order. Eighteen minutes from folder to live landing page with a working Resend email integration. The post walks through the exact prompts, the .claude/plans/ save behavior that's new, and when each mode is the right one.

Read Episode 2 →

Anthropic doubled Claude Code's limits – and signed for SpaceX's Colossus 1 data center

Anthropic doubled Claude Code limits and signed for SpaceX Colossus 1 data center

If you use Claude Code on a Pro, Max, Team, or seat-based Enterprise plan, your 5-hour rate limits just doubled. The peak-hours throttle on Pro and Max is gone too. Both effective immediately.

The reason both happened on the same day: Anthropic signed an agreement with SpaceX to use the entire compute capacity of SpaceX's Colossus 1 data center. Over 300 megawatts. More than 220,000 NVIDIA GPUs. Online inside one month. This is the most consequential infrastructure announcement Anthropic has made in 2026 and it tells you exactly what the next 12 months of Claude Code will be priced like.

Read the limits + SpaceX breakdown →

Claude Code 2.1.131 – the silent 1-hour prompt cache bug, plus 64 changes since 2.1.126

Claude Code 2.1.131 cover - 1-hour prompt cache TTL silently downgrading to 5 minutes

If you set cache_control: { ttl: "1h" } anywhere in your Claude Code prompts and have been quietly paying for the 1-hour cache while wondering why your hit rate looked off – 2.1.129 just fixed it. The TTL was silently downgrading to 5 minutes. Nobody could see it. The dashboards looked fine. The bills weren't.

That's the buried headline of this release pair. 2.1.131 on May 6 ships two surgical fixes (a VS Code-on-Windows activation bug and a Mantle endpoint auth header). 2.1.129 the same day shipped 27 entries. 2.1.128 on May 4 shipped 35. Sixty-four individual changes since 2.1.126 on May 1, and the most expensive one was invisible. The post unpacks the cache bug in detail plus the 4 most useful changes from the 64.

Read the 2.1.131 recap →

Every WotAI live session is now public on YouTube

Every WotAI live session now public on YouTube - n8n, Claude Code, MCP, AI agents

The full WotAI live session archive is now public on the WotAI YouTube channel – every Monday/Wednesday/Friday call, every workflow walk-through, every Claude Code release breakdown. Free, no signup, no wall.

If you've wanted to see how production-grade n8n automations get built (without sitting through another "what is n8n" intro), or you wanted a real engineer's take on the last 20 Claude Code releases, the back catalog is yours. The 759 members of the WotAI Skool community join three live calls a week; now you can watch them all.

Browse the archive →


Reading order if you want the through-line: start with Episode 2 (how to actually use Claude Code's four modes), then the SpaceX/limits piece (why Anthropic just doubled what you can do with them), then 2.1.131 (the cache bug to fix today), and finish with the recordings (how we use this stuff in production three times a week).

Reply and tell me which one moved a needle for you. I read every reply.

– Alex


P.S. – Production-Grade Claude Code in 5 Days, June 1 cohort. Same setup I used to ship the Episode 2 build, unpacked across 5 days on your own codebase – with me there to fix what doesn't translate. Early bird is $297 through May 15 (saves $200 off the $497 regular price).

Save my seat →

Reply to this email anytime – I read every one.

WotAI • wotai.co

WotAI · AI Automation that Actually Ships

wotai.co

WotAI · 440 N. Barranca Ave., Unit 3777, Covina, CA 91723

WotAI

If you're building with Claude Code, n8n, or AI agents in production, this is the email I'd want. Weekly recap of what shipped, what broke, and how I'm running it. Specific over vague. Replies welcome.

Read more from WotAI
The Opus 4.8 number nobody's posting

The Opus 4.8 number nobody's posting Not the coding score. Plus Open Design at 4 weeks, and 2 more reads. The Opus 4.8 number nobody's posting Hey Reader, Anthropic shipped Claude Opus 4.8 this week, and the whole feed posted the same screenshot – the coding score. That's the least interesting part. The number that actually changes how you work is below, plus three more reads stacked behind it. Claude Opus 4.8 – the number nobody's posting Everyone screenshotted the coding score (64.3% to...

What goes underneath Claude Claude for Small Business is a plugin, not a tier. Plus 3 more reads. What goes underneath Claude Hey Reader, Anthropic shipped Claude for Small Business on May 13. A week in, most people still think it's a new subscription tier parallel to Pro and Team. It isn't. It's a plugin. And whether you want it depends on what you're actually trying to get Claude to do. Four reads below. Claude for Small Business – what's actually in the plugin Anthropic shipped Claude for...

Should AI write your code unsupervised? My answer + today's source-control post. Plus 2.1.141 and Episode 2. Should AI write your code unsupervised? Hey Reader, Short answer to the subject line: yes – but only with the workflow underneath. "Unsupervised" doesn't mean nobody's reviewing, it means the human moves from typing to reviewing. The three reads below are the layers of that workflow stack. Plan Mode is how you organize a session before the agent runs. Source control is the safety net...