[BUG] 2.1.9 is unusable

Resolved 💬 8 comments Opened Jan 16, 2026 by heyJordanParker Closed Feb 28, 2026

Preflight Checklist

  • [x] I have searched existing issues and this hasn't been reported yet
  • [x] This is a single bug report (please file separate reports for different bugs)
  • [x] I am using the latest version of Claude Code

What's Wrong?

I'm on macOS 26.2 (25C56) using Ghostty (latest) and using Claude Code inside tmux.

Performance on long-running tasks has always been unreasonably bad. In general, it's not uncommon for CC to use over 10 GIGS of RAM for 3-4 instances if they're running subagents.

But today, with 2.1.9, it's simply unusable.

I'm seeing:

  1. CC instances completely freezing
  2. CC instances NOT being saved to history
  3. The flickering bug is worse than ever & appears on shorter conversations

One of my CC sessions hit 30 gigs of ram usage (I'm at 64, conveniently) and 200% CPU usage.
This is absolutely ridiculous and it's 100% coming from the harness rendering, not the actual IO operations or the tiny amount of actual data CC can operate with at any given time.

Oh, I have the verbose flag on. That shouldn't matter, but I'll validate; the rendering engine is finicky so it's worth testing.

What Should Happen?

I would expect:
<1GB of ram usage for a CC session, no matter how subagents I run in it
<30% CPU usage

No scrolling bug.
No blowing up my machine because of the TUI.

Error Messages/Logs

Steps to Reproduce

Run a code review of a branch with 8 parallel subagents.

Claude Model

Opus

Is this a regression?

Yes, this worked in a previous version

Last Working Version

2.1.8

Claude Code Version

2.1.9

Platform

Anthropic API

Operating System

macOS

Terminal/Shell

Other

Additional Information

_No response_

View original on GitHub ↗

This issue has 8 comments on GitHub. Read the full discussion on GitHub ↗