Everyone on the team is on Claude now. This is the difference between two people getting the same result — one for a fraction of the tokens. Short modules, a one-page cheat sheet, then a graded quiz.
A one-page summary now that you've been through the modules. Each line is explained in full above — use this as a refresher before the quiz.
Cost = tokens × model's per-token rate. The whole conversation resends every turn — a long stale session costs more even for a one-line question.
Default Sonnet. Haiku for high-volume/simple. Opus/Fable only for genuinely hard reasoning.
Default medium/high. Drop to low for simple or high-volume work. Reserve xhigh/max for frontier problems.
Disable unused ones via /mcp. Prefer a plain CLI (gh, aws, gcloud) when one exists.
Check /context before guessing. Move rarely-used CLAUDE.md instructions into skills.
Name the file/line, not just the symptom. Cut "triple-check everything"-style rituals from standing prompts.
Put the document(s) first, your question last — measurably better answers, fewer costly re-asks.
Wrap distinct content in XML tags (<context>, <instructions>) so Claude doesn't misparse and re-ask.
Ask for a targeted lookup, not "explore the codebase" — vague asks trigger broad, expensive scanning.
Keep stable content first, dynamic content last. Don't touch the system prompt or tool list mid-session. Cache dies after 5 min–1 hr idle.
/clear for new work, /compact mid-task. /rename before clearing. /rewind to undo a bad turn instead of arguing it back.
Delegate verbose/exploratory work. Pin cheap ones to Haiku via /agents.
Give a long agentic task an advisory token budget so it self-regulates instead of running unbounded.
~7× the tokens of a normal session. Keep teams small, shut teammates down once done.
Put shared context in the Project once, not pasted per chat. Keep project knowledge lean.
A self-check, not a test — answer each question, then see which ones you got right. Retake it as many times as you like.