[Anthropic: Resolved Claude Code Performance Degradation Issue and Reset Usage Limits for All Subscribers] Odaily Planet Daily News - According to official sources, Anthropic previously received feedback regarding a decline in the quality of Claude Code, which was caused by three independent changes involving Claude Code, Claude Agent SDK, and Claude Cowork. The specific issues included: a reduction in default inference intensity from high to medium on March 4, which led to a decrease in intelligence levels—this setting was restored on April 7; a bug in the caching optimization feature launched on March 26, which mistakenly cleared old inference records, causing the model to exhibit forgetting and repetition—this was fixed on April 10; and system prompt words added on April 16 to reduce redundancy, which negatively impacted programming quality—this was rolled back on April 20. All affected models have now returned to normal, and Anthropic has reset the usage limits for all subscribers.