Do I need to pay for anything?
No. The core kit runs on free Claude and free Notion - and every claim on these pages was verified on exactly those plans. Costs only appear if you later want specific add-ons - the scheduled briefings add-on needs a paid Claude plan, and each add-on guide says so upfront. Nothing in the core nudges you toward paying.
What are the different paid plan options?
The short version: Free is enough for the core kit. If you find yourself wanting more, the step most people take is Pro.
Free - no card, no commitment. Message limits apply, but they're not stingy for conversational use. The kit's core setup (memory pages, weekly dumps, "where am I at?" conversations) fits comfortably. The one thing you can't do on Free is scheduled briefings, which need Claude to run automatically on a timer - that's a paid feature.
Pro - Anthropic's individual paid tier. Higher message limits (including Opus, which is capped on Free), and the ability to use Claude on a schedule. If you're using the kit daily and hitting limits, or if you want the morning briefing add-on, Pro is the upgrade. Most people who go paid stop here.
Max (5x) - five times the usage of Pro. For heavy users who burn through Pro limits regularly.
Max (20x) - twenty times Pro's limits. For people whose work genuinely demands it - or who've tried 5x and still need more.
For current prices: claude.ai/upgrade has the live comparison.
Is my stuff private?
Your memory pages live in your own Notion account, in plain view - you can read every word Claude writes. Your conversations with Claude are covered by Anthropic's privacy policy, and the pages by Notion's. Practical advice: don't put things in the system you wouldn't put in either service, and if you ever want out, everything exports (see below).
What exactly can Claude see in my Notion?
It connects with your access level to the workspace you approved - Notion's consent screen puts it as "respect your page and database access and permissions" and "take actions on your behalf, based on your access level." In practice the kit has it work only in its own memory pages, and everything it does is visible. If you want a harder wall between Claude and an existing Notion life, make a free second Notion account just for the kit - Guide 2 covers it.
What are these "Claude wants to use..." permission pop-ups?
Claude asking to use its own hands. The first time it uses each Notion ability (creating pages, updating them, searching), a card appears - choose Always allow each time. There are a few, then they stop. Denying them breaks the quiet-saving the kit depends on.
Claude says it can't reach Notion. What do I do?
Disconnect and reconnect Notion in Settings → Connectors, then ask again. If it ever asks you to log in to Notion again out of the blue, that's the connection needing a refresh - same fix.
Can I use it just on my phone?
Yes - that's rather the point. The project, the conversations, and the Notion pages are all cloud-side, so your phone sees the same relationship your computer does. The copy-and-paste steps during setup are slightly easier at a keyboard; everything after works anywhere.
Which model should I pick? The menu confused me.
Don't overthink it - the default is the right answer. The model picker in the message box matters for some kinds of work, but for the kit's job (remembering you, sorting your dumps, answering "where am I at?") the default model does it well. That default is Sonnet - capable, quick, and what most people use most of the time. Pick nothing, change nothing, move on.
If you're curious what the models are: Anthropic offers a few tiers. Haiku is the lightest and fastest - good for quick, simple questions. Sonnet is the default and the sweet spot for most things: it reasons well, writes well, and responds quickly. Opus is the most powerful - slower and uses more of your plan's allowance, but better for genuinely complex tasks like long-form analysis, difficult writing, or working through something tangled. For the kit's day-to-day work, Sonnet is the right call. You might reach for Opus occasionally for something heavy - but you'd know when.
And here's the quietly reassuring part: because your Claude's knowledge of you lives in its memory pages rather than in any particular model, model changes barely touch the relationship. When Anthropic releases a new model or retires an old one, a kit Claude reads its own pages and still knows you - your patterns, your list, how to speak to you. The voice might shift a little; the knowing doesn't. (If a new model's tone feels off, say so - corrections land in its pages and carry forward, same as always.)
I already use Notion / Todoist / a paper diary. Does this replace them?
No. The kit is a memory, not a method. Keep whatever works; Claude's pages sit alongside. (If you're a heavy Notion user, you might enjoy that the memory lives somewhere you already look - or you might prefer the second-account wall. Both fine.)
What happens if I stop using it for ages?
Nothing. No streaks, no guilt, no backlog shame. Come back whenever, ask "where am I at?", and it picks up. You don't even need to end sessions properly - wandering off mid-conversation is a fully supported exit. The kit was built by someone whose weeks get eaten sometimes; it's designed for the return, not the streak.
Do I have to have ADHD?
No. It was built for ADHD first, and the design choices (no-shame language, quick wins first, structured check-ins instead of "how are you feeling?") come from that. But an external memory is useful to most brains with more life than working memory - other neurodivergencies, carers, anyone in a season of overload.
What if I want to leave?
Everything is yours and everything exports. Notion pages export to markdown (plain text files that open anywhere), and your project instructions are just text you can copy. Nothing in the kit locks you in - walking away costs nothing but the walk.
What are the add-ons?
Optional extensions, each making the system one notch more proactive: Notifications (tasks ping your phone via Todoist), Briefings (a morning briefing before you wake and an evening roundup, on a schedule - paid Claude plan), and Own your files (your memory as plain markdown on your own machine instead of Notion). Guides coming; the kit is complete without them.
Something else?
The kit's repo is public, and Held in Time is where the thinking lives - both good places to look, and I'm reachable through either.
