#skills

Create a skill that forces a red→green verification loop so Claude cannot claim done without running the suite.

Every installed skill spends context just being listed, before you invoke anything. Sort the list by cost, then set the ones you never use to name-only.

End implementation prompts with /simplify so parallel agents review reuse, efficiency, quality, and CLAUDE.md compliance in one sweep.

A skill that checks a link or reformats a file does not need your session model. Set model, effort and context: fork in the frontmatter so cheap skills stay cheap.

Build slash skills that accept issue numbers, branch names, or formats via $ARGUMENTS / $0 so one command covers many cases.

Shift specialized playbooks into on-demand skills so every session does not pay for instructions you rarely need.

When a skill hits an edge case, append it to a gotchas section in that skill rather than to CLAUDE.md. The lesson then loads only when the skill runs.

Mark deploy/ship/destructive skills so only an explicit /command runs them—Claude cannot auto-trigger them mid-session.

Save repeatable workflows as SKILL.md (or .claude/commands) so you slash-invoke consistent PR reviews, releases, or checklists.

"Never add comments" in CLAUDE.md plus "make this readable" in your prompt is a contradiction Claude has to resolve before it writes a line. Find those pairs and pick one.

Most CLAUDE.md lines exist to correct a model that no longer makes that mistake. Delete them after an upgrade and re-add only the ones a failure proves you need.