Fetching from the wire…
01
02
03
04
05
06
07
08
Public story · 2026-07-25 · high
Max plan users get the new default automatically, and any harness that never pinned a specific model ID just started running a different model overnight.
Why now: Version 2.1.219 shipped July 24, and anyone running Claude Code on a schedule needs to check which model their next run will actually call.
Claude Code shipped v2.1.219 on July 24, and it makes claude-opus-5 the default Opus model with a 1M-token context window, per the release notes. Opus 4.7 drops out of fast mode entirely, so /fast now only applies to Opus 5 and Opus 4.8.
Pricing holds at $5 per million input tokens and $25 per million output, same as Opus 4.8. Fast mode still costs double the base price for roughly 2.5x speed. Nothing changed there.
What changed is who gets the new default without asking for it. Max plan users get claude-opus-5 automatically. If your harness never pins a specific model ID, it's now calling a different model than it was calling yesterday. You didn't touch your code, and your output changed anyway.
I've hit this failure mode before. A harness with no pinned model works fine for months. Then it starts producing different output, and you spend an afternoon convinced you broke something in your own code. The fix is boring: pin the model ID.
The release notes name this directly as the likely explanation when a harness changes behavior with no code changes on your end. That's a specific admission, not a vague warning.
Each link below shares sources, entities, or timing with this story.
Within five days of each other, both Claude Code (v2.1.158, May 31) and Cursor (3.6, May 29) shipped remarkably similar architectures for autonomous agent execution. Both use a classifier subagent that reviews each pending action against conversation context and decides: allow...
One line in the v2.1.215 changelog, July 19: Claude Code no longer invokes the /verify and /code-review skills on its own. You call them explicitly now (changelog). If your workflow assumed a review pass fired at the end of a task, it doesn't anymore, and nothing told you. Cod...
Two researchers in my set surfaced this independently, which is usually a sign it matters. Claude Code 2.1.166, first seen June 6, introduces a fallback-models setting: configure up to three models tried in order when the primary is overloaded or unavailable. It also adds glob...
Two releases in two days. v2.1.78 adds effort, maxTurns, and disallowedTools frontmatter to agent definition files — per-agent behavior tuning without code changes. New StopFailure hook fires on API errors. v2.1.77 raised default output to 64K tokens and the upper bound to 128...
Two releases landed today and they're both significant. The headline security fix: PreToolUse hooks returning allow were bypassing deny permission rules — including enterprise managed settings. If you're running Claude Code in a managed enterprise environment with security hoo...
Two things in one release, and only one of them is in the security section. CLAUDE_CODE_SUBAGENT_MODEL used to override the model for every subagent. In 2.1.251 it sets the default instead. An agent definition's model: frontmatter now wins, and so does an explicit per-spawn mo...
MindPattern daily
One email a day at 7 AM. Sources and a take on every story. Unsubscribe anytime.