Plan To Win: AI and Coaching Disclosure (with Coaching Addendum)

1. About This Disclosure

1.1 This AI and Coaching Disclosure (the "Disclosure") explains how the artificial-intelligence features of Plan To Win and the Map of You work, what data they use, what they cannot do, how optional founder coaching works, and how the serious-safety notification works. It is one of the four Legal Documents you accept before you start using the app or the Map of You, and it forms part of our Terms of Service. "We", "us" and "PTW" mean Zackery Frank Boyle, carrying on business as Plan To Win. "You" means the person using the Service.

1.2 Defined terms. "AI Features" means the parts of the Service that use an AI model to generate output, including the Map of You's question and reading generation. "AI Output" means what they generate. "Coaching" means optional 1-on-1 coaching with the founder. "Coaching Addendum" means Section 9.

2. What the AI Features Do

We keep this list current with each release and check it before we publish a new version.

• Coach chat and the AI Assistant. You ask questions and it answers. It can see your schedule, add tasks, propose plans you accept or decline, and look things up on the web for facts such as exercise technique or nutrition.
• Plans, goals and days. Goal generation, clarifying and refining a goal, checking whether a plan is realistic, fitness activity generation, side quests (custom, refined or suggested), and Shuffle (rearranging your day in your own words).
• Reflection ("Go deeper", "Map your values", identity work and skill-tree branches). Short guided questions about something on your mind, which produce a reading of your values, your identity and an "insight" that you can keep or let go.
• The Map of You. It writes each quiz question, analyses your answer to build your values map, and writes your reading (archetype, values, growth edge, gap and first moves).
• Reading documents. Importing a schedule, brief or plan you upload (PDF, Word or text) into projects and tasks when the simple line-by-line reader is not enough, and pulling tasks out of a project brief. The text is read; the file is not kept (Privacy Policy, Section 5.1).
• Food. Estimating a meal from a photo or a description (a draft you confirm or discard).
• Suggestions. Essential App suggestions for the moment (using the time of day, your next task and the apps you chose), Vision Board and Bucket List ideas.
• Coach memory and look-back. Facts you tell the coach (after it asks), and, only if you allow it, your own journal, mood and activity when you ask it to look back (Sections 4.6 and 4.7).
• Plans the coach builds. Build my day or week, and a nightly plan for tomorrow, both shown for your approval before anything is added (Section 4.8).
• Tool Studio. An AI writes a small tool for you from your instructions (Section 3.9).
• Reading a shared screenshot. Text or a screenshot you share to Plan To Win and choose to send to the coach (Section 3.8).
• Coaching tool (clients only). Drafting check-in messages for your coach to review and send personally (Section 9.5).
• The AI writes in the language you choose in Settings.

3. How AI Processing Works

3.1 The data flow. When you use an AI feature, our server (not your phone) sends the relevant message, plus the background described in Section 4, to Anthropic, which runs the model and sends back an answer. If you use a photo or file with a feature (a meal photo, a chat attachment, a schedule or plan document), it is sent too: a chat photo is used for that reply only and not stored; a meal photo is stored only if you save the meal; a document's text is extracted and sent, and the file itself is not stored.

3.2 Web search. The AI can use a web search tool run by our AI provider. The words it searches for may be sent to a search service.

3.3 Anthropic is our service provider. It processes your data on our behalf under a data-processing agreement. As Anthropic publishes: it does not train its models on the content submitted through its API; it automatically deletes API inputs and outputs from its own systems within 30 days of receipt, except where it retains data longer to investigate suspected misuse of its terms (up to two years) or where the law requires it; and it may process data in the United States. We rely on Anthropic's published terms for these facts. We use two Anthropic models: a larger one for the coach, plans and readings, and a smaller one for quick tasks such as meal estimates and app suggestions. To keep replies fast, a short-lived cached copy of our shared instructions is held by Anthropic for minutes at a time, within the same deletion terms.

3.4 We do not train AI models on your data. We do not use your content, journal, reflections, quiz answers or chat text to train AI models, and we do not sell it.

3.5 Limits. We limit how much AI you can use in a month, and we count usage per account. Limits depend on your plan and are shown in Settings > Membership. A message that trips our care screen is never blocked by a limit.

3.6 Storage. The full text of your AI Assistant conversations is saved in your account so you can look back at it, and you can delete a conversation in the app. A message that trips our care screen (Section 8) is not saved.

3.7 Voice. If you talk to the coach, your browser or phone turns your speech into text, and depending on your device that speech recognition may be done by Google or Apple under their own privacy terms. We do not receive or store the audio, only the text you send. The coach's replies are read aloud with a voice already on your device.

3.8 Screenshots and shared text (Send to PTW). Text or a screenshot you share from another app stays on your phone and is shown to you first, and you can edit it, hide numbers, emails and links, or cut off parts of a screenshot. It is sent to the AI only when you tap Ask the coach. A shared screenshot is used for that one reply and is not stored or kept in your coach history. Shared messages can name other people; only what you keep is used, and the coach is told to use only a first name and what your plan needs, and never to save other people's phone numbers or addresses.

3.9 Tool Studio. When you ask for a tool, your instructions are sent to the AI, which writes the tool's code. The result is checked by our code and runs in a sandbox with no internet access, and with only the permissions you allow. AI-written code can contain mistakes, so you are responsible for how you use a tool you build (Terms, Section 12.10).

3.10 Plans, habits and Life setup. When you build a plan for a goal, our instructions to the AI include rules for small daily habits. If you use Life setup to describe your day, what you tell it is sent to the AI to draft a day for you to review. Nothing is added until you accept it. The questions-only path in Life setup does not use the AI.

4. What the Coach Can See

4.1 Always. The coach chat always sees: your phase, XP, streak, which pillars are active, today's tasks, your fixed commitments for today and tomorrow (including any imported from Google Calendar), plans you accepted from the coach, a summary of your lifting progress when you ask about training, any plan you added from a health provider (for example a physio plan), the Essential Apps you chose, and your onboarding "coaching profile" answers (your guidance style, biggest obstacle, motivation style, north star habit and pillar satisfaction ratings).

4.2 You choose the rest. In Settings > AI Coach Profile there are six switches:

4.3 Never shown to the coach, whatever your switches: the raw answers you type inside a "Go deeper" session, anything that triggered the care screen (Section 8), which is not saved, and anything a linked person shared with you in a social feature.

4.4 Changing your mind. You can change these switches at any time. Changes apply to future conversations. Text already sent to the AI provider in the past is covered by the Privacy Policy.

4.5 Consent. These choices are why we ask for your consent before you start using the app or the Map of You. See the Privacy Policy for how we treat sensitive information.

4.6 Memory. The coach can remember short facts you tell it about yourself and a short note about each finished chat, and uses them quietly in later chats. It asks "Remember this?" before it keeps a lasting fact, never keeps an address, phone number or email address, and keeps nothing that tripped the care screen. You can see and delete everything it remembers under Chat history > What I remember about you, or tell the coach to forget something. The switch Settings > Plan & coach > Coach memory stops anything new being remembered and stops saved memory being used. What you say now and today's data always count for more than something remembered.

4.7 Looking back at your journal, mood and activity. This is its own switch, Look back when you ask (Settings > Plan & coach > AI Coach Profile), and it is off by default; the coach asks the first time it would need it. When it is on, the coach can read your own journal entries, mood check-ins and activity to answer a question such as "how has my week been?", and its reply shows an "I looked at" receipt of which part of your data it read and the dates it covered. What it read is sent to Anthropic for that reply and is not saved to memory unless you tell the coach to remember it. Entries that sound heavy are not handed over for analysis; the coach answers gently and points to real help. It does not diagnose, and this is not medical advice. The raw answers you type inside a Go deeper session and anything that tripped the care screen are still never included.

4.8 Changes the coach makes. When the coach proposes a change, such as Build my day, a nightly plan for tomorrow or an edit to a task, you see it as a card and nothing is applied until you approve it (unless you have switched on "Apply by itself" for the nightly plan). You can Undo what the coach did for 7 days. With Let the coach make changes turned off in Settings > Plan & coach, the coach can still read your day and talk with you but cannot change anything.

5. Accuracy and Limits

5.1 What the AI is not. The AI is not a therapist, counsellor, doctor, nurse, dietitian, lawyer, accountant or financial adviser. It does not diagnose, name conditions, or predict how your life will go.

5.2 It can be wrong. It can be wrong, out of date, biased or overconfident, and it may misunderstand you. It is designed to be warm and to point you to real help when something is heavy, but it can miss things. AI Output can differ each time you ask. A plan it builds from a document you upload can misread a date or a detail; review it before you accept it.

5.3 Your final call. Everything is your final call. Check what matters with a qualified person before you act on it. See the Health and Wellness Disclaimer.

5.4 Do not rely on it in an emergency. If you are in immediate danger, contact your local emergency number.

6. What We Can See

6.1 Admin spot-check log. Our admin panel shows the founder short excerpts (the first 500 characters) of your messages to the AI Assistant and goal-refinement chat and its replies, together with your name and email, so we can spot bad or broken answers. It is a spot-check log, not a full transcript. Messages that trip the care screen are never logged. These excerpts are kept for 90 days and then deleted.

6.2 Other things the founder can see. In the admin panel the founder can see your account and profile details (such as your name, email, date of birth, height, weight and Valueprint reading), your task names and activity numbers, your AI Coach Profile answers, error logs, bug reports you send, notes we keep about your account for support, and an estimate of what your AI use costs us. If someone reports you, or you report someone, in a social feature, the founder sees the report, both names and emails, and a copy of the reported text (Privacy Policy, Section 8.4). The admin panel does not show your journal or reflection text.

6.3 What we do not do. We do not use your content to train AI models ourselves, and we do not sell it.

7. Your Choices

• Turn off what the coach can see (Section 4.2), what it remembers and whether it can look back at your journal, mood and activity (Sections 4.6 and 4.7), and whether it can make changes (Section 4.8).
• Delete an AI conversation in the app, or delete your account (Privacy Policy, Section 14).
• Download your data (Settings > Account & Security > Download my data).
• Choose not to use the AI Features. The Map of You needs the AI to generate your reading; the rest of the Service still works without it. Some tools open an outside AI assistant in your browser with a prompt you can paste; Plan To Win sends nothing in that case.

8. The Serious-Safety Notification

8.1 What we do. The app has a gentle safety layer. When something you write in a reflection or in the AI chat looks like signs of serious danger (for example self-harm, thoughts of suicide, or danger happening right now):

• the app stops the exercise and shows you a kind message, an encouragement to reach out to someone you trust, the local emergency number message, and a link to findahelpline.com to find a free, confidential crisis line in your country;
• the AI chat is told to respond with warmth and without tools, tasks or plans in that moment;
• what you wrote in that moment is not saved (in reflection) and is not logged in our admin panel.

8.2 The notification. If our systems detect signs you may be in serious danger, we may also notify the PTW team. The notice contains your account and the time only, not the text you wrote, so that a person can reach out to you (for example by email) to check in and point you to real help.

8.3 The detail. For completeness: the notice also records which part of the app it came from (for example the AI chat or a guided reflection). It is sent to the founder by email and appears in the admin panel, where the founder can mark it handled with a private note. We send at most one notice per person in any 24 hours, and only if you have accepted the current consent screen including Box 3. When a notice was actually sent, the app tells you so on screen, and it does not say so otherwise.

8.4 What we do not promise.

• We cannot monitor in real time. Nobody is watching the app.
• We do not promise the notice will be seen, or seen quickly, or that anyone will reach out. We promise no response time.
• We are not an emergency service, and we do not contact emergency services for you.
• The detection can miss things and can also be wrong (it can react to a phrase you meant in another way). A missing or unexpected notice or message is not a judgment about you.
• The safety layer is not a substitute for real help.

8.5 What you should do. If you are in immediate danger, contact your local emergency number now (911 in Canada and the US, 112 in the EU, 999 in the UK). Find a crisis line at findahelpline.com. Talk to someone you trust today. You can also email us at hello@plantowin.app, but do not rely on email in an emergency.

8.6 Who sees the notice and what happens to it. The notice goes to the PTW team (the founder and any admin the founder authorises). It is kept for 12 months after it is marked handled, and is used only to reach out and to improve safety. It is not sold or shared, except where we are legally required to, or where we believe it is necessary to prevent serious harm to someone.

8.7 If you have not accepted. If you have not accepted the current Legal Documents, we do not send a notice, and the crisis screen keeps its plain wording.

8.8 Reports about someone else. If a person you are linked with in a social feature reports that they are worried about your safety, the report goes to the founder as urgent and the person reporting is shown crisis resources. We do not promise any action or timing (Section 8.4).

9. Coaching Addendum (Optional 1-on-1 Coaching)

This Addendum applies only if you sign up for coaching with the founder. It is in addition to the Terms of Service.

9.1 What coaching is. Founder-led coaching walks you through the same loop the app runs: your Valueprint, your Horizon, your pillars, a daily plan, weekly review and adjustments. Coaching is offered as one or more packages, which may include a Map session, a template-building session, weekly or periodic 1-on-1 calls, messaging between calls, a shared plan document, and an intake form before the first session. The packages on offer, their length, and their price are described to you when you book.

9.2 What coaching is not. Coaching is not therapy, counselling, medical care, licensed clinical care or treatment. Your coach does not diagnose or treat any condition and does not give medical advice. If something heavy comes up, your coach will name it gently and encourage you to see a professional.

9.3 No guaranteed outcomes. We do not promise any result: not fitness, weight, income, relationship or life results. How far you get depends on you. You stay in charge: everything is your final call.

9.4 Sharing your progress with your coach. If you are a coaching client, and only after your coaching consent is confirmed, your coach can see your progress in the app: your profile and reading, your goals and tasks, your logged activity and weight, your mood check-ins, and the excerpts of your AI Assistant chats. The app shows "Your progress is shared with your coach" when this applies. Your coach does not see anything you have not agreed to share. You can ask to stop sharing at any time by emailing hello@plantowin.app or telling your coach, and coaching may change or end if we cannot see enough to coach you.

9.5 Notes, records and drafted messages. The coach's private notes, call schedule and agreed action items are stored in our coaching tool. Notes are kept for as long as you are a client and for a reasonable period after, to give continuity if you return. You can ask what is in them. To help your coach write a check-in message to you between calls, our coaching tool may send a short summary of the facts above (your first name, streak, completion, open action items, your coach's recent notes and, with your consent confirmed, your recent mood check-ins and activity logs, never your journal or reflections) to our AI provider to draft the text. Your coach reviews and sends every message personally.

9.6 Calls and recording. Calls may be held on Zoom, Google Meet or similar. Calls are recorded only with your permission, and you can say no. A recording, if made, is kept only as long as needed for coaching and then deleted.

9.7 Confidentiality. What you tell your coach stays between you and us, and we use it only to coach you and to improve the product in ways you have agreed to. This confidentiality has limits: we may disclose information where the law requires it, or where we reasonably believe it is necessary to prevent serious harm to someone. Your data and your story are yours. If we want to use a testimonial or your story, we will ask first, and any published testimonial will say if you received free access or coaching from us.

9.8 Your responsibilities. Be honest with your coach about relevant health issues. Get your doctor's approval before you start physical activity or diet changes. Let us know if you cannot make a call as early as you can. Treat your coach with respect.

9.9 Fees, cancellation and refunds. Fees, and how to cancel or reschedule, are stated to you when you book, and are paid through Stripe. Nothing here limits refund rights you have under consumer-protection law.

9.10 Ending coaching. Either of us can end coaching. At the end, we hand you off to the app, and your coach's access stops unless you keep sharing.

10. Changes and Contact

We may update this Disclosure, and we will ask you to accept again if the change is material (Terms, Section 3). Questions: hello@plantowin.app · 4718 Stanley Rd SW, Calgary, Alberta T2S 2R2, Canada.