Claude Code adds an agent environment checkup

Boris Cherny says Claude Code now has a /checkup command that cleans unused skills, MCPs, and plugins, deduplicates local and checked-in CLAUDE.md files, splits root instructions into nested files and skills, disables slow hooks, updates Claude Code, enables auto mode, and pre-approves common read-only commands with confirmation before changes. Why it matters: Agent performance increasingly depends on environment hygiene: context bloat, duplicated instructions, stale tools, slow hooks, and noisy permission prompts all degrade real work. A built-in checkup turns maintenance of the harness into a first-class workflow, not an occasional manual cleanup.

Try this: Run the same idea as a checklist before scaling an agent setup: inventory unused skills/MCPs/plugins, dedupe local versus repo instructions, move broad root guidance into scoped nested files, measure hook latency, and only pre-approve low-risk read-only commands after reviewing the proposed changes.