[Bug] Poor readability of checkpoint feedback and degraded output quality at 40-70% context window utilization

Bug Description
Feedback on Claude Code (CC) Issues: Frequent interruptions requiring user decisions, combined with poor readability of updates/feedback Claude Code frequently pauses mid-task to check in with me and request decisions. However, the information and feedback provided at these checkpoints often has poor readability, making it very difficult for me to keep up with CC's progress (or significantly increasing the cognitive/decision-making cost on my end). As a result, this substantially undermines the efficiency and quality of collaborative work with CC, rather than improving it. Noticeable degradation in work quality when context window usage is between 40%–70% The quality of Claude Code's output noticeably declines once the context window utilization falls within the 40%–70% range.

Environment Info

• Platform: linux
• Terminal: windows-terminal
• Version: 2.1.226
• Feedback ID: 41064b3c-49c7-413a-b431-b331ec671d92

Errors

[]