Turn a Messy Plan Into a Shared Board in Claude Whiteboards
Talk a messy plan into a Claude whiteboard, fix what Claude laid out by hand and by asking, then invite your team to the board.
Check your Claude Pro usage, then run a task on fewer messages: Low effort, unused apps off, questions before Claude builds, edits and a clean handoff.
Rundown Pro
Every course and recording, the certification program and member perks, for $29 a month. Guides stay free.
You'll check your Claude Pro usage, set up a chat that spends less of it, and run one real task with Nate Grahek's sketch-first prompt before handing it to a fresh chat. As of November 2026, Anthropic publishes no message count for Pro, and these habits stretch the five-hour and weekly limits without removing them. Anthropic's new Claude experience is rolling out gradually, so your web search and Edit controls may differ from these screens.
Step 1 of 7
Find out how much of your current five-hour session and your week you've used, and when each resets. You'll come back here later to see whether the habits help.
Pro has two limits running at once. The session limit starts with your first message and resets five hours later; the weekly limit resets at a time Anthropic assigns to your account. Anthropic's help page names the session bar Current session. This guide was recorded on a higher plan, so the exact label of the weekly bar on Pro couldn't be checked.
Step 2 of 7
Set the new chat to Low effort before you send anything. Anthropic says Low and Medium work well for routine tasks and stretch your usage further.
The menu itself says higher effort takes longer and uses your limits faster, and on Sonnet 5.5 the Max level is tagged 5× or more usage. On Sonnet 5.5 and Opus 5.5, thinking is always on, so effort is the setting that decides how much Claude thinks. Low is enough for planning, short questions and the sketch in step 4.
Step 3 of 7
Turn off any connected apps this chat won't use, and keep Claude from searching the web when the task doesn't need current information. Anthropic says tools and connectors are token-intensive, and web searches count toward your limits.
No-search line
Don't search the web for this. Use only what I've given you.In Anthropic's new Claude experience there's no web search toggle, and Claude searches when it decides it helps. Anthropic suggests asking Claude not to search when you don't need current information, which is what the line above does. If your message box still shows Chat and Cowork options, you have the previous experience, and the Web search item in the + menu turns search on and off.
Step 4 of 7
Start the task with a prompt that tells Claude to ask before it builds, answer all its questions in one reply, and get a short sketch back with nothing built yet. The tip is Nate Grahek's, from Nate's Notebook in The Rundown on 5 Nov 2026 ("Make AI sketch before it builds"), written for people on $20 plans.
Sketch-first prompt (Nate Grahek's tip)
Don't start yet. I want to [your task]. Conserve your energy until we agree on the path. Ask me up to five smart multiple-choice questions about what I want, and keep every reply to a paragraph or two. After I answer, sketch the plan in the chat and wait for my go-ahead.
Each time Claude builds the wrong thing, you spend a message correcting it and get another long reply back. A few questions and a short sketch catch the mistake before anything is built. Claude decides when to show clickable inputs (multiple choice, multi-select or a ranking) below the conversation, and you can always type instead. For more on the technique, see Billy Howell's guide on getting better AI outputs with multiple-choice questions, Nate's workshop recording on context engineering (chapters 5 and 6), or the lesson "Let it ask before it drafts" in the course Turn a task into an AI workflow.
Step 5 of 7
Reply to the sketch once, with your go-ahead, an answer to each open question, every change and any question of your own. Anthropic's own advice is to group related questions and send them together.
Go-ahead template
Go. [Your answer to each question Claude left open.] Also change: [every change, all here]. [Any question you have.] Keep your reply to [what you want back].One-message task template
Goal: [what you want made]. It's for: [who will use or read it]. What I have: [notes, files, numbers]. My questions: [all of them]. Keep the answer to [length].For the Ember & Wick sketch, a filled-in go-ahead would answer both decisions (cart behavior, and whether to say the shop is fictional) and list any section changes in the same message. For writing tasks, Anthropic suggests laying out your requirements, audience and key points up front, which is what the task template asks for.
Step 6 of 7
When a reply goes wrong because of something you sent, fix that message and save it. Claude answers again from there, and the wrong reply drops out of the chat.
The (i) icon in the edit box says: "Saving sends this as your latest message and restarts from here." Everything after the edited message is replaced by Claude's new answer. That's why editing the latest message is usually the right choice, and editing an early one starts that part of the chat over.
Step 7 of 7
When a chat has run long, have Claude sum it up in under 150 words and paste that into a new chat. Anthropic suggests starting a new conversation when you're near your limit in a long chat.
Handoff prompt
This chat is getting long. Write a handoff summary under 150 words that I can paste as the first message of a new chat: the goal, what we've decided, and what's still open. No code.End the summary with your rules for replies, so the new chat follows them from its first message. In the recorded run the last line read "Keep replies to a paragraph or two. Don't write code until I approve." The lesson "Start fresh with a handoff summary" in Turn a task into an AI workflow covers the idea in more depth, and the course Your first AI workflow goes further.
Mark it complete to keep your progress. The background and common questions follow.
These Rundown Pro recordings cover the same tool or topic as this guide.
$29/mo, cancel anytime
There's no set number. As of November 2026 Anthropic publishes no message count for Pro and says it may adjust limits. How fast you use your allowance depends on how long and complex your chats are, the features you use, the model and the effort level. The session limit starts with your first message and refills five hours later; the weekly limit resets at a fixed time assigned to your account.
claude.ai, the desktop app and Claude Code share one allowance. If you mostly use Claude Code, see our guide on using fewer Fable tokens with an orchestrator setup.
Wait for the reset time shown in Settings > Usage, turn on usage credits on the same page (billed at standard API rates, separately from your $20 subscription), or upgrade. Members in the Claude Duet research preview (US only, through a waitlist) also get three limit resets on Pro, used from Settings > Usage; unused ones expire on 15 Dec 2026.
Only by paying with usage credits. Pro's monthly price doesn't include it, while Max includes Fable up to half the weekly limit. Sonnet 5.5 and Opus 5.5 are in the model menu on Pro.
Anthropic says project content is cached and counts less against your limits when you reuse it, though the cache expires after a period of inactivity. Setting up a project is outside this guide.
Anthropic's new Claude experience, which merges chat and Cowork, is rolling out gradually to Pro, Max and Ultra. If your message box still shows Chat and Cowork options, you don't have it yet, so you may see a Web search item and other labels. Once you have it you can't switch back, and Anthropic says usage may be measured slightly differently during the rollout. The screens here were recorded on 15 Nov 2026 on a higher plan, mostly in incognito chats, so the Usage page and model list may differ slightly from yours.
The sketch-first prompt, the one-message templates and the handoff summary work in any chat assistant. The menus in this guide are Claude's. As of November 2026, OpenAI's Plus page says Plus may include message caps that vary with demand. OpenAI's 6 Aug 2026 release note gave unlimited text chats to Free and Go users. On Plus, Work and Codex share one allowance that can have a five-hour and a weekly window, shown in Settings, and OpenAI says it isn't a fixed number of messages.
Check your connection, then try again.