kalei/docs/product/app-blueprint.md

# [APP NAME TBD] — Complete Product Blueprint

### Science-Backed Manifestation & Thought Transformation App
**Version:** 1.0 — February 2026
**Author:** Matt (concept & research) + Claude (architecture & strategy)
**Status:** Pre-development — Ready for implementation

---

## Table of Contents

1. [Executive Summary](#1-executive-summary)
2. [The Problem & Opportunity](#2-the-problem--opportunity)
3. [Scientific Foundation](#3-scientific-foundation)
4. [The 6 Steps to Manifestation](#4-the-6-steps-to-manifestation)
5. [Product Vision & Core Features](#5-product-vision--core-features)
6. [User Experience Flows](#6-user-experience-flows)
7. [AI Strategy & Prompt Architecture](#7-ai-strategy--prompt-architecture)
8. [Technical Architecture](#8-technical-architecture)
9. [Database Schema](#9-database-schema)
10. [Monetization Model](#10-monetization-model)
11. [Cost Estimates](#11-cost-estimates)
12. [Implementation Roadmap](#12-implementation-roadmap)
13. [Building with Claude Code & Codex](#13-building-with-claude-code--codex)
14. [App Store Strategy](#14-app-store-strategy)
15. [Growth & Future Features](#15-growth--future-features)
16. [Naming Candidates](#16-naming-candidates)
17. [Research References](#17-research-references)
18. [Appendix: AI Prompt Templates](#18-appendix-ai-prompt-templates)

---

## 1. Executive Summary

This app is a science-backed personal transformation tool that does two things no competitor does well together:

1. **Transforms negative thoughts on the spot** using AI personalized to the user's preferred coaching style — brutal honesty, gentle guidance, logical analysis, philosophical perspective, or humor.
2. **Guides users through a structured 6-step manifestation system** grounded in peer-reviewed research from goal-setting theory, implementation intentions, self-efficacy, selective attention, habit formation, and network effects.

Unlike the vast majority of manifestation apps (which lean on "Law of Attraction" magical thinking), this app is built on published, peer-reviewed psychology and neuroscience research. The critical differentiator is Step 3: "Believe it's possible — but not guaranteed." This aligns with Bandura's self-efficacy theory (capability belief, not outcome certainty) and avoids the toxic positivity trap that causes most people to abandon manifestation when results aren't immediate.

**Target users:** Adults 18-45 interested in personal growth, goal achievement, and mental wellness — but skeptical of woo-woo self-help. The "I want to improve my life, but I want science, not crystals" market.

**Revenue model:** Freemium with premium subscription ($8/mo or $60/yr).

**Tech stack:** React Native + Expo (frontend) → Supabase (backend/database/auth) → Claude API (AI engine).

---

## 2. The Problem & Opportunity

### The Problem

The self-help/manifestation app space is dominated by two extremes:

1. **Pure woo-woo apps** — Vision boards, affirmations, "raise your vibration" content. No scientific backing. Users get a dopamine hit, then quit when nothing changes because the app never connected belief to *action*.

2. **Clinical CBT/therapy apps** — Evidence-based but clinical, cold, and expensive. Feel like homework. Users who aren't in acute distress don't stick with them.

There's a massive gap in the middle: people who want to actively shape their lives using *proven* methods, delivered in a way that feels personal, warm, and actionable — not clinical or magical.

### The Opportunity

- The meditation/mindfulness app market (Calm, Headspace) proved people will pay monthly for daily mental wellness tools. But meditation isn't for everyone — many people want *active* thought work, not passive stillness.
- AI makes personalized coaching scalable for the first time. What previously required a $200/hour therapist or coach (personalized thought reframing, goal scaffolding) can now happen instantly on a phone.
- The "manifestation" keyword has massive search volume but almost zero credible competition. Most results are embarrassingly unscientific. An app that owns "science-backed manifestation" has a blue ocean.

### Competitive Landscape

| App | Category | Weakness vs. Our App |
|---|---|---|
| Reframe (alcohol reduction) | Behavioral change | Focused solely on drinking, not general thought transformation |
| Reframe AI (mental health) | CBT-based reframing | Generic AI coaches, no manifestation/goal system |
| Headspace / Calm | Meditation | Passive, not active thought work |
| Manifest (various) | Affirmations/vision boards | Zero science, pure "Law of Attraction" |
| BetterHelp / Talkspace | Therapy | Expensive ($240+/mo), requires scheduling, clinical positioning |

**Our moat:** The combination of (1) AI-powered personalized reframing with (2) a structured, research-cited goal achievement system, delivered with (3) anti-toxic-positivity guardrails and transparent science.

---

## 3. Scientific Foundation

Every feature in this app maps to published, peer-reviewed research. This isn't decoration — it's the product's structural integrity.

### Research Pillars

**Goal Setting & Clarity**
- Locke & Latham (2002) — *Building a Practically Useful Theory of Goal Setting and Task Motivation: A 35-Year Odyssey*
- Locke, Latham, et al. — *New Directions in Goal-Setting Theory*
- Core insight: Specific, challenging goals with clear commitment lead to higher performance than vague "do your best" goals. This drives our Step 1 (Decide/Clarity) and the AI-guided SMART goal creation process.

**Implementation Intentions (Planning & Action)**
- Gollwitzer (1999) — *Implementation Intentions: Strong Effects of Simple Plans*
- Gollwitzer (2006) — *Implementation Intentions and Goal Attainment*
- Core insight: "If-then" plans ("If situation X arises, then I will do Y") dramatically increase follow-through on goals. This is the backbone of our Step 5 (Act in Alignment) and every micro-action the AI generates.

**Visualization & Mental Practice**
- Schuster et al. (2011) — *Best Practice for Motor Imagery: A Systematic Literature Review*
- Liu et al. (2025) — *Effects of Imagery Practice on Athletes' Performance*
- Seok et al. (2023) — *Mental Practice in Stroke Rehabilitation*
- Core insight: Mental rehearsal improves real-world performance across domains. Drives our Step 2 (See It) with personalized visualization scripts.

**Self-Efficacy (Belief → Behavior)**
- Bandura (1977) — *Self-Efficacy: Toward a Unifying Theory of Behavioral Change*
- Core insight: Belief in one's *capability* (not guaranteed outcomes) is the strongest predictor of goal achievement. This shapes our Step 3 (Believe It's Possible — But Not Guaranteed), which is the app's philosophical soul. We explicitly reject certainty-based belief ("the universe will provide") in favor of capability-based belief ("I have or can develop the skills to make this happen").

**Attention & Selective Processing**
- Yantis (2008) — *The Neural Basis of Selective Attention*
- Stevens & Bavelier (2012) — *Attention, Learning, and Brain Plasticity*
- Koch & Tsuchiya (2007) — *Attention vs Consciousness*
- Core insight: What you pay attention to shapes your reality — not metaphysically, but neurologically. Your brain filters millions of inputs daily, and attention training can literally change what you perceive. This drives our Step 4 (Notice Differently) and the Reframer's pattern analysis.

**Habits & Repetition**
- Wood, Mazar & Neal (2021) — *Habits and Goals in Human Behavior*
- Wood & Neal (2007) — *A New Look at Habits and the Habit-Goal Interface*
- Core insight: ~43% of daily behavior is habitual, not deliberate. Sustained change requires building automatic routines, not just willpower. Drives our Step 6 (Repeat and Compound) and streak tracking.

**Network Effects & Opportunity**
- Granovetter (1973) — *The Strength of Weak Ties*
- Core insight: Transformative opportunities come disproportionately from weak social ties, not close friends. Drives future community features where anonymous sharing creates unexpected connections and inspiration.

**Expectancy Effects & Placebo**
- Pardo-Cabello et al. (2022) — *Placebo/Nocebo Effects: Mechanisms & Evidence*
- Stetler et al. (2013) — *Expectations & Placebo Response*
- Core insight: Expectations directly influence physiological and behavioral outcomes. Supports the overall premise that structured positive expectation (when combined with action) measurably improves outcomes.

---

## 4. The 6 Steps to Manifestation

These steps form the app's conceptual spine. Every feature maps back to them.

```
┌─────────────────────────────────────────────────────────┐
│                                                         │
│  Step 1: DECIDE (Clarity)                               │
│  ─────────────────────────                              │
│  Define exactly what you want and why it matters.       │
│  Research: Locke & Latham goal-setting theory            │
│                                                         │
│  Step 2: SEE IT (Mental Rehearsal)                      │
│  ─────────────────────────────────                      │
│  Visualize the process and outcome with specificity.    │
│  Research: Schuster et al., Liu et al. motor imagery    │
│                                                         │
│  Step 3: BELIEVE IT'S POSSIBLE (But Not Guaranteed)     │
│  ──────────────────────────────────────────────────     │
│  Build capability-based confidence, not blind faith.    │
│  Research: Bandura self-efficacy theory                  │
│                                                         │
│  Step 4: NOTICE DIFFERENTLY (Attention)                 │
│  ──────────────────────────────────────                 │
│  Train your brain to filter for aligned evidence.       │
│  Research: Yantis, Stevens & Bavelier selective attn     │
│                                                         │
│  Step 5: ACT IN ALIGNMENT (Daily)                       │
│  ────────────────────────────────                       │
│  Execute micro-actions using if-then planning.          │
│  Research: Gollwitzer implementation intentions           │
│                                                         │
│  Step 6: REPEAT AND COMPOUND                            │
│  ───────────────────────────                            │
│  Build automatic habits that sustain momentum.          │
│  Research: Wood & Neal habit formation                    │
│                                                         │
└─────────────────────────────────────────────────────────┘
```

### Why Step 3 Is the App's Soul

Most manifestation content says: "Believe with absolute certainty and the universe will deliver."

This app says: "Believe you're *capable* of making this happen — but understand it's not guaranteed, and that's okay. Certainty isn't the point. Aligned action is."

This single distinction:
- Prevents the toxic positivity trap (and subsequent crash when things don't work)
- Aligns with Bandura's actual research (self-efficacy is about capability, not outcomes)
- Builds resilience when setbacks happen (because they were expected as possible)
- Makes the app credible to the skeptical-but-curious demographic that most manifestation apps repel

---

## 5. Product Vision & Core Features

The app has two core pillars that work independently but compound together.

### Pillar 1: The Reframer (Daily Thought Transformer)

This is the app's sticky daily-use feature and primary acquisition tool. It's what users come for before they even know about the Manifestation Engine.

**What it does:** User inputs a negative thought or experience → AI generates a personalized reframe in their chosen coaching style → ends with a micro-action (if-then implementation intention).

**Reframe Styles (user selects primary + secondary during onboarding):**

| Style | Tone | Example Response |
|---|---|---|
| **Brutal Honesty** | Direct, no-nonsense | "You're catastrophizing. Here's what's actually true..." |
| **Gentle Coach** | Warm, empathetic | "That sounds really hard. Let's find what you can take from this..." |
| **Logical Analyst** | Structured, evidence-based | "Let's break this down. What evidence supports vs contradicts this thought?" |
| **Philosophical** | Big-picture, perspective-shifting | "Consider: what would this look like from 10 years in the future?" |
| **Humor** | Light, disarming | "Okay, worst case scenario... you become a cautionary tale in a self-help book. Best case?" |

**Reframe Structure (every response follows this pattern):**
1. **Acknowledge** — Never dismiss the feeling. Validate first.
2. **Reframe** — Apply the research-backed perspective shift. When appropriate, cite the science: "Your brain is doing what Yantis's research on selective attention predicts — you're filtering for the negative."
3. **Micro-action** — End with a Gollwitzer-style if-then intention: "If [situation] happens again, then I will [specific response]."

**Sub-features:**
- **Daily prompt** — Push notification at user-set time: "What's weighing on you today?"
- **Reframe history & pattern analysis** — Over time, the app surfaces patterns: "You tend to catastrophize about work on Mondays" or "Your negative thoughts cluster around relationships." This is Step 4 (Notice Differently) applied to the user's own thought life.
- **"Flip It" quick mode** — One-tap mode for acute stress. Voice-record or quick-type → instant reframe. Zero friction.

### Pillar 2: The Manifestation Engine (Goal System)

This is the premium feature. A structured journey through all 6 steps, guided by AI.

**Step 1 — Decide (Clarity):**
- AI guides user through Locke & Latham's SMART framework via conversational prompts
- Makes it feel personal, not corporate ("What would change in your life if you achieved this?" not "Define a measurable objective")
- Output: Crystal-clear goal statement + personal "why it matters" statement
- Free tier: Users can create ONE goal (enough to get invested, creates cliffhanger for premium)

**Step 2 — See It (Mental Rehearsal):**
- AI generates a personalized visualization script based on the user's goal
- Text format initially; optional audio in later phases (text-to-speech)
- Based on Schuster et al.'s motor imagery best practices: first-person, sensory-rich, process-focused (not just outcome)
- User sees this as a "Vision Summary" they can revisit daily

**Step 3 — Believe It's Possible:**
- AI walks user through identifying self-efficacy barriers
- Prompts: "What makes you doubt this is possible?" → systematically addresses each doubt
- Sources of evidence: past successes, transferable skills, role models, small wins
- Generates a "Belief Statement" — a personalized, evidence-based declaration of capability
- Anti-toxic-positivity: Never says "you definitely will." Says "you have real reasons to believe you *can*."

**Step 4 — Notice Differently:**
- Daily "attention prompts" — "Today, notice one thing that's already aligned with your goal"
- Trains selective attention (Yantis, Stevens & Bavelier) toward goal-aligned evidence
- Users log observations → builds an "Evidence Journal" that reinforces belief over time
- Over time, AI surfaces patterns: "You've noticed 23 instances of alignment this month. Your brain is rewiring."

**Step 5 — Act in Alignment:**
- AI generates daily micro-actions using Gollwitzer's if-then structure
- Format: "If it's [time/situation], then I will [specific 15-minute action toward goal]"
- Users check off completed actions
- Streak tracking with smart reminders
- AI adapts difficulty: starts tiny (2-minute actions), scales up as habit solidifies

**Step 6 — Repeat & Compound:**
- Habit tracking dashboard showing compound progress over time
- Visual growth charts, milestone celebrations
- AI periodically revisits and updates the plan based on progress
- Weekly AI-generated summary: reframes, patterns, goal progress, suggested adjustments
- Celebrates the *process* (consistency) not just outcomes

### Additional Features

**Anti-Manifestation Guard:**
When the AI detects a user spiraling into toxic positivity, magical thinking, or unhealthy attachment to outcomes, it gently course-corrects:
> "Manifestation isn't about wishing — it's about aligning your attention, beliefs, and actions with what you want. Let's focus on what you can control today."

**Science Drawer:**
Each step has an expandable "Why This Works" section citing the actual research. Example:
> "This technique is based on Gollwitzer's 1999 research on Implementation Intentions, which showed that people who create if-then plans are 2-3x more likely to follow through on goals than those who simply set intentions."

This builds trust and differentiates from every competitor that says "trust the process" without explaining why.

**Weekly Reflection:**
AI-generated weekly summary covering:
- Reframe count and patterns
- Mood trajectory (inferred from thought inputs)
- Goal progress against daily actions
- Suggested adjustments for the coming week

**Streak & Gamification (Light):**
- Daily streaks for reframing and goal actions
- "Reframe count" milestone celebrations (10, 50, 100, 500...)
- Level-ups tied to *consistency* (not volume — doing it daily matters more than doing 10 in one day)
- No leaderboards, no competitive mechanics. This is personal transformation, not a game.

### Connector Features (Science-Driven Engagement)

These three features weave the core pillars into a cohesive daily practice. Each directly implements a research pillar that was underserved by the core feature set.

**The Rehearsal (Guided Visualization — Lens sub-feature):**
Step 2 (See It) implemented as a concrete feature. AI-generated, personalized mental rehearsal scripts tied to the user's active Lens goals. Following Schuster et al.'s best-practice protocol and Liu et al.'s optimal dosage (~10 min, 3x/week), each Rehearsal walks users through the *process* of achieving their goal in first-person, present-tense, multi-sensory detail — including obstacle rehearsal to prevent the toxic positivity trap. Always paired with a follow-up micro-action. All visuals are SVG-generated (progress rings, completion patterns).

**The Ritual (Context-Anchored Daily Flow):**
Step 6 (Repeat and Compound) implemented as active habit scaffolding, not just passive streak tracking. A timed, sequenced daily flow (Morning/Evening/Quick templates) that chains 2-3 Kalei activities into one context-anchored routine. Directly implements Wood & Neal's (2007) core finding that context stability predicts habit formation. The user builds *one habit* ("my morning Ritual") instead of three separate ones. The Ritual tracks context consistency (same time, same sequence) per Wood et al. (2021), not just frequency. Addresses the week 3-4 retention cliff by making daily usage automatic before novelty fades. UI is a single flowing screen with SVG progress segments — no navigation, no friction.

**The Evidence Wall (Mastery Tracking — You tab):**
Upgrades Kalei's self-efficacy building from Bandura's weakest source (verbal persuasion via reframes) to his *strongest* source (mastery experiences). Automatically collects proof of capability from all features: completed micro-actions, saved keepsakes, self-correction moments, streak milestones, goal completions, reframe echoes in later writing. The AI surfaces this evidence contextually when it detects self-efficacy dips — "Here's what I've seen: you've completed 23 micro-actions this month." Visual design is a growing SVG mosaic of color-coded geometric proof-point tiles. Evidence, not cheerleading.

---

## 6. User Experience Flows

### Flow 1: First Open (Onboarding)

```
1. Welcome screen
   "Transform your thinking. Achieve your goals. Backed by science."
   [Get Started]

2. How this works (3 swipeable cards)
   Card 1: "Reframe your negative thoughts daily with AI"
   Card 2: "Build a science-backed plan for what you want in life"
   Card 3: "Watch your thinking patterns transform over time"

3. Choose your reframe style
   "When life gets hard, how do you prefer to hear the truth?"
   [Show 5 style cards with examples — user taps primary, then secondary]

4. Set your daily check-in time
   "When should we ask how you're doing?"
   [Time picker — defaults to 8:00 PM]

5. Create account
   [Email / Google / Apple Sign-In]
   
6. First reframe
   "Let's try it right now. What's been on your mind?"
   [Text input → AI reframe → "wow" moment]

7. Home screen
```

### Flow 2: Daily Reframe

```
1. Push notification: "What's weighing on you today?"
   [User taps → opens app to input screen]

2. Input screen
   Text field: "What happened? What's bothering you?"
   [Optional: voice input button]
   [Submit]

3. Processing (1-3 seconds)
   Brief animation (not a spinner — something warm)

4. Reframe card
   ┌─────────────────────────────────────┐
   │ Your thought:                       │
   │ "I bombed the presentation and      │
   │  everyone thinks I'm incompetent"   │
   │                                     │
   │ [Brutal Honesty] reframe:           │
   │                                     │
   │ "Let's be real: you're not a mind   │
   │ reader. You have zero evidence that │
   │ 'everyone' thinks anything. What    │
   │ you DO know is one presentation     │
   │ didn't go as planned..."            │
   │                                     │
   │ 💡 Your micro-action:               │
   │ "If I start replaying the           │
   │ presentation, then I will write     │
   │ down 3 things that DID go right."   │
   │                                     │
   │ 📚 Why this works ▼                 │
   │                                     │
   │ [Save] [Share] [Try Different Style] │
   └─────────────────────────────────────┘

5. Save → returns to home with updated streak
```

### Flow 3: Manifestation Engine (Goal Creation Wizard)

```
1. "Create a Manifestation" [+]

2. Step 1: Decide
   AI conversation (3-5 exchanges):
   "What do you want to achieve?"
   → "Why does this matter to you personally?"
   → "What would your life look like if this happened?"
   → "Let's make this specific..." [SMART refinement]
   → Output: Clarity Statement displayed as a card

3. Step 2: See It
   AI generates visualization script
   "Close your eyes and imagine..."
   [User reads/listens → marks as complete]

4. Step 3: Believe
   AI conversation:
   "What makes you doubt this is possible?"
   → [User lists doubts]
   → AI addresses each with evidence + past successes
   → Output: Belief Statement card

5. Step 4: Notice
   AI sets up attention prompts
   "Starting tomorrow, I'll send you one prompt per day"
   [Preview first prompt]

6. Step 5: Act
   AI generates first week of if-then micro-actions
   [User reviews and confirms]

7. Step 6: Compound
   Dashboard appears showing the full system:
   - Goal card
   - Today's attention prompt
   - Today's micro-action
   - Streak counter
   - Progress timeline

8. "Your manifestation engine is running. Show up daily."
```

### Flow 4: Weekly Summary

```
Push notification (Sunday evening):
"Your weekly transformation report is ready ✨"

┌─────────────────────────────────────┐
│ WEEK 3 SUMMARY                      │
│                                     │
│ 🔄 Reframes this week: 6           │
│    Pattern: Work stress (Mon/Tue)   │
│    Shift: Catastrophizing → present │
│                                     │
│ 🎯 Goal: [user's manifestation]    │
│    Actions completed: 5/7           │
│    Evidence logged: 4 observations  │
│                                     │
│ 📈 Trend: Your negative thoughts    │
│    are 23% less catastrophic than   │
│    Week 1. Your brain is learning.  │
│                                     │
│ 💡 Suggestion for next week:        │
│    "Your Tuesday stress pattern     │
│    suggests a preparation ritual    │
│    Monday night could help..."      │
│                                     │
│ [View Full Report] [Share Progress] │
└─────────────────────────────────────┘
```

---

## 7. AI Strategy & Prompt Architecture

### Why Claude API

Claude is the right AI engine for this app because:
1. **Tone control** — Claude excels at adopting specific communication styles while maintaining safety guardrails
2. **Nuance** — Thought reframing requires acknowledging complexity, not defaulting to toxic positivity. Claude handles nuance better than most models.
3. **Safety** — Claude's constitutional AI approach means it won't encourage harmful thinking patterns, which is critical for a mental wellness app
4. **Cost** — Claude Haiku (for simple reframes) and Sonnet (for complex goal guidance) provide a cost-effective tier structure

### Model Selection by Feature

| Feature | Model | Rationale | Est. Cost/Call |
|---|---|---|---|
| Simple reframe | Claude Haiku 4.5 | Fast, cheap, good enough for single-thought reframes | ~$0.003 |
| Complex reframe (with patterns) | Claude Sonnet 4.5 | Needs to analyze history + generate nuanced response | ~$0.008 |
| Goal creation wizard | Claude Sonnet 4.5 | Multi-turn conversation requiring coherent guidance | ~$0.015 |
| Visualization script | Claude Sonnet 4.5 | Creative, sensory-rich, personalized writing | ~$0.012 |
| Weekly summary | Claude Sonnet 4.5 | Pattern analysis across multiple data points | ~$0.020 |
| Pattern detection | Claude Haiku 4.5 | Categorization task, doesn't need deep reasoning | ~$0.003 |

### System Prompt Architecture

All AI calls go through Supabase Edge Functions (never direct from client). Each function includes:

1. **System prompt** — Defines the AI's role, constraints, and output format
2. **User context** — Injected from the database (reframe style, goal data, history)
3. **Research grounding** — Relevant citations pre-loaded so the AI can reference them naturally
4. **Anti-toxicity guardrails** — Explicit instructions to avoid toxic positivity, magical thinking, and dismissal of real problems

See [Appendix: AI Prompt Templates](#18-appendix-ai-prompt-templates) for full prompt examples.

### Rate Limiting & Cost Control

- Free tier: 2 reframes/day (hard limit enforced at Edge Function level)
- Premium tier: Unlimited (soft limit of 20/day with warning, hard limit of 50/day)
- All responses cached for 24 hours — if user submits identical thought, return cached reframe
- Goal creation wizard: Max 10 AI exchanges per goal setup
- Weekly summaries: Generated once per week via cron job, not on-demand

---

## 8. Technical Architecture

### Stack Overview

| Layer | Technology | Why This Choice |
|---|---|---|
| **Frontend** | React Native + Expo | Single codebase for iOS, Android, web. JavaScript/TypeScript (most transferable skill). Expo handles 90% of native complexity. Huge community + AI coding tool support. |
| **Backend/DB** | Supabase | Managed Postgres DB, built-in auth (email, Google, Apple Sign-In), Edge Functions (Deno runtime), file storage, real-time subscriptions. No servers to manage. |
| **AI Engine** | Claude API (Anthropic) | Called via Supabase Edge Functions. Never direct from client (cost control, rate limiting, prompt security). |
| **Push Notifications** | Expo Push Notifications | Free, built into Expo, works on both platforms out of the box. |
| **Payments** | RevenueCat | Handles App Store + Google Play subscriptions. Way easier than native in-app purchase implementation. Manages receipts, trials, and entitlements. |
| **App Builds** | Expo EAS | Cloud builds for iOS and Android. Handles code signing, certificates, TestFlight, and OTA updates. |
| **Analytics** | PostHog (or Mixpanel free tier) | Open-source product analytics. Event tracking, funnels, retention analysis. |

### Architecture Diagram

```
┌──────────────────────────────────────────────────────────────┐
│                    User's Phone / Web Browser                │
│                  (React Native + Expo App)                   │
│                                                              │
│  ┌─────────┐ ┌──────────┐ ┌──────────┐ ┌────────────────┐  │
│  │Reframer │ │Manifest. │ │ History  │ │    Settings     │  │
│  │  Screen │ │  Engine  │ │ & Stats  │ │  & Profile      │  │
│  └────┬────┘ └────┬─────┘ └────┬─────┘ └───────┬────────┘  │
│       │           │            │                │            │
│       └───────────┴────────────┴────────────────┘            │
│                           │                                  │
│              Supabase JS Client SDK                          │
└──────────────────────────┬───────────────────────────────────┘
                           │ HTTPS
                           ▼
┌──────────────────────────────────────────────────────────────┐
│                        Supabase                              │
│                                                              │
│  ┌──────────────────────────────────────────────────────┐   │
│  │  Auth (email, Google Sign-In, Apple Sign-In)         │   │
│  │  - JWT tokens, session management                    │   │
│  │  - Row Level Security (RLS) on all tables            │   │
│  └──────────────────────────────────────────────────────┘   │
│                                                              │
│  ┌──────────────────────────────────────────────────────┐   │
│  │  PostgreSQL Database                                  │   │
│  │  - profiles (user preferences, reframe style)        │   │
│  │  - reframes (thought log + AI responses)             │   │
│  │  - manifestations (goals, clarity statements)        │   │
│  │  - daily_actions (if-then micro-actions)             │   │
│  │  - attention_logs (evidence journal)                 │   │
│  │  - streaks (habit tracking)                          │   │
│  │  - subscriptions (premium status via RevenueCat)     │   │
│  └──────────────────────────────────────────────────────┘   │
│                                                              │
│  ┌──────────────────────────────────────────────────────┐   │
│  │  Edge Functions (Deno TypeScript)                     │   │
│  │  - POST /reframe → validates, calls Claude, saves    │   │
│  │  - POST /generate-plan → goal wizard AI calls        │   │
│  │  - POST /generate-visualization → Step 2 scripts     │   │
│  │  - POST /weekly-summary → pattern analysis + report  │   │
│  │  - POST /check-subscription → RevenueCat validation  │   │
│  │  - All functions enforce rate limits & auth           │   │
│  └──────────────────────────────────────────────────────┘   │
│                                                              │
│  ┌──────────────────────────────────────────────────────┐   │
│  │  Storage                                              │   │
│  │  - Profile pictures                                   │   │
│  │  - Audio visualization scripts (Phase 2+)            │   │
│  │  - Exported reports (PDF generation, Phase 3+)       │   │
│  └──────────────────────────────────────────────────────┘   │
│                                                              │
│  ┌──────────────────────────────────────────────────────┐   │
│  │  Cron Jobs (pg_cron or Supabase scheduled functions) │   │
│  │  - Weekly summary generation (Sundays 6pm user TZ)   │   │
│  │  - Streak reset check (daily midnight user TZ)       │   │
│  │  - Pattern analysis batch job (weekly)                │   │
│  └──────────────────────────────────────────────────────┘   │
└──────────────────────────┬───────────────────────────────────┘
                           │ HTTPS (from Edge Functions only)
                           ▼
┌──────────────────────────────────────────────────────────────┐
│                 Claude API (Anthropic)                        │
│                                                              │
│  - Thought reframing (Haiku for simple, Sonnet for complex) │
│  - Goal guidance & conversational wizards                    │
│  - Visualization script generation                           │
│  - Pattern analysis & weekly summaries                       │
│  - Anti-toxicity guardrails baked into system prompts        │
└──────────────────────────────────────────────────────────────┘

┌──────────────────────────────────────────────────────────────┐
│              External Services                                │
│                                                              │
│  RevenueCat ──── Subscription management, receipt validation │
│  Expo EAS ────── Cloud builds, OTA updates, push notifs      │
│  PostHog ─────── Product analytics, funnel tracking          │
└──────────────────────────────────────────────────────────────┘
```

### Key Architecture Decisions

**Why Supabase Edge Functions for AI calls (not direct from client):**
1. API key security — never expose the Claude API key in the client bundle
2. Rate limiting — enforce free/premium limits server-side
3. Prompt security — system prompts stay on the server (can't be extracted)
4. Cost control — can add caching, deduplication, and usage tracking
5. Response formatting — can sanitize/validate AI responses before sending to client

**Why Expo over bare React Native:**
1. Managed workflow — no need to touch Xcode or Android Studio for most tasks
2. EAS Build — cloud builds mean Matt doesn't need a Mac for iOS builds
3. OTA updates — push bug fixes without App Store review
4. Built-in push notifications — no Firebase setup needed
5. Web support — same codebase runs on web via Expo Web

**Why RevenueCat over native IAP:**
1. Handles the nightmare of App Store receipt validation
2. Single API for both iOS and Android subscriptions
3. Built-in analytics (MRR, churn, trial conversion)
4. Webhook support for server-side subscription status
5. Free up to $2,500/mo MTR

---

## 9. Database Schema

### Core Tables

```sql
-- ==========================================
-- PROFILES
-- User preferences and settings
-- ==========================================
CREATE TABLE profiles (
    id UUID PRIMARY KEY DEFAULT gen_random_uuid(),
    user_id UUID REFERENCES auth.users(id) ON DELETE CASCADE NOT NULL UNIQUE,
    display_name TEXT,
    reframe_style_primary TEXT NOT NULL DEFAULT 'gentle_coach'
        CHECK (reframe_style_primary IN (
            'brutal_honesty', 'gentle_coach', 'logical_analyst', 
            'philosophical', 'humor'
        )),
    reframe_style_secondary TEXT
        CHECK (reframe_style_secondary IN (
            'brutal_honesty', 'gentle_coach', 'logical_analyst', 
            'philosophical', 'humor'
        )),
    notification_time TIME DEFAULT '20:00:00',
    timezone TEXT DEFAULT 'UTC',
    onboarding_completed BOOLEAN DEFAULT FALSE,
    subscription_tier TEXT DEFAULT 'free' 
        CHECK (subscription_tier IN ('free', 'premium')),
    reframes_today INTEGER DEFAULT 0,
    last_reframe_date DATE,
    created_at TIMESTAMPTZ DEFAULT NOW(),
    updated_at TIMESTAMPTZ DEFAULT NOW()
);

-- ==========================================
-- REFRAMES
-- Thought transformation history
-- ==========================================
CREATE TABLE reframes (
    id UUID PRIMARY KEY DEFAULT gen_random_uuid(),
    user_id UUID REFERENCES auth.users(id) ON DELETE CASCADE NOT NULL,
    original_thought TEXT NOT NULL,
    reframed_response TEXT NOT NULL,
    reframe_style_used TEXT NOT NULL,
    category TEXT, -- AI-inferred: 'work', 'relationships', 'health', 'finance', 'self-worth', 'general'
    sentiment_score DECIMAL(3,2), -- -1.00 to 1.00 (negative to positive)
    if_then_action TEXT, -- The micro-action from the reframe
    is_quick_flip BOOLEAN DEFAULT FALSE, -- Was this a "Flip It" quick reframe?
    created_at TIMESTAMPTZ DEFAULT NOW()
);

-- Index for history queries and pattern analysis
CREATE INDEX idx_reframes_user_date ON reframes(user_id, created_at DESC);
CREATE INDEX idx_reframes_user_category ON reframes(user_id, category);

-- ==========================================
-- MANIFESTATIONS
-- Goal/manifestation tracking (the 6-step engine)
-- ==========================================
CREATE TABLE manifestations (
    id UUID PRIMARY KEY DEFAULT gen_random_uuid(),
    user_id UUID REFERENCES auth.users(id) ON DELETE CASCADE NOT NULL,
    title TEXT NOT NULL,
    -- Step 1: Decide
    clarity_statement TEXT,
    why_it_matters TEXT,
    -- Step 2: See It
    visualization_text TEXT,
    visualization_audio_url TEXT, -- Future: TTS audio file
    -- Step 3: Believe
    belief_statement TEXT,
    doubts_addressed JSONB, -- Array of {doubt, response} objects
    -- Metadata
    status TEXT DEFAULT 'active' 
        CHECK (status IN ('active', 'paused', 'completed', 'archived')),
    current_step INTEGER DEFAULT 1 CHECK (current_step BETWEEN 1 AND 6),
    created_at TIMESTAMPTZ DEFAULT NOW(),
    updated_at TIMESTAMPTZ DEFAULT NOW()
);

CREATE INDEX idx_manifestations_user ON manifestations(user_id, status);

-- ==========================================
-- DAILY ACTIONS
-- If-then micro-actions (Step 5)
-- ==========================================
CREATE TABLE daily_actions (
    id UUID PRIMARY KEY DEFAULT gen_random_uuid(),
    user_id UUID REFERENCES auth.users(id) ON DELETE CASCADE NOT NULL,
    manifestation_id UUID REFERENCES manifestations(id) ON DELETE CASCADE NOT NULL,
    if_trigger TEXT NOT NULL, -- "If it's 7am..."
    then_action TEXT NOT NULL, -- "...then I will spend 15 min writing"
    scheduled_date DATE NOT NULL,
    completed BOOLEAN DEFAULT FALSE,
    completed_at TIMESTAMPTZ,
    created_at TIMESTAMPTZ DEFAULT NOW()
);

CREATE INDEX idx_daily_actions_user_date ON daily_actions(user_id, scheduled_date);

-- ==========================================
-- ATTENTION LOGS
-- Evidence journal entries (Step 4)
-- ==========================================
CREATE TABLE attention_logs (
    id UUID PRIMARY KEY DEFAULT gen_random_uuid(),
    user_id UUID REFERENCES auth.users(id) ON DELETE CASCADE NOT NULL,
    manifestation_id UUID REFERENCES manifestations(id) ON DELETE CASCADE NOT NULL,
    prompt_text TEXT, -- The attention prompt that was shown
    observation TEXT NOT NULL, -- What the user noticed
    created_at TIMESTAMPTZ DEFAULT NOW()
);

CREATE INDEX idx_attention_logs_user ON attention_logs(user_id, created_at DESC);

-- ==========================================
-- STREAKS
-- Habit tracking data
-- ==========================================
CREATE TABLE streaks (
    id UUID PRIMARY KEY DEFAULT gen_random_uuid(),
    user_id UUID REFERENCES auth.users(id) ON DELETE CASCADE NOT NULL,
    streak_type TEXT NOT NULL 
        CHECK (streak_type IN ('reframe', 'daily_action', 'attention_log', 'overall')),
    current_count INTEGER DEFAULT 0,
    longest_count INTEGER DEFAULT 0,
    last_activity_date DATE,
    created_at TIMESTAMPTZ DEFAULT NOW(),
    updated_at TIMESTAMPTZ DEFAULT NOW(),
    UNIQUE(user_id, streak_type)
);

-- ==========================================
-- WEEKLY SUMMARIES
-- Pre-generated weekly reports
-- ==========================================
CREATE TABLE weekly_summaries (
    id UUID PRIMARY KEY DEFAULT gen_random_uuid(),
    user_id UUID REFERENCES auth.users(id) ON DELETE CASCADE NOT NULL,
    week_start DATE NOT NULL,
    week_end DATE NOT NULL,
    summary_text TEXT NOT NULL,
    reframe_count INTEGER DEFAULT 0,
    action_completion_rate DECIMAL(5,2),
    attention_log_count INTEGER DEFAULT 0,
    top_category TEXT,
    mood_trend TEXT, -- 'improving', 'stable', 'declining'
    created_at TIMESTAMPTZ DEFAULT NOW()
);

CREATE INDEX idx_weekly_summaries_user ON weekly_summaries(user_id, week_start DESC);

-- ==========================================
-- ROW LEVEL SECURITY
-- Users can only access their own data
-- ==========================================
ALTER TABLE profiles ENABLE ROW LEVEL SECURITY;
ALTER TABLE reframes ENABLE ROW LEVEL SECURITY;
ALTER TABLE manifestations ENABLE ROW LEVEL SECURITY;
ALTER TABLE daily_actions ENABLE ROW LEVEL SECURITY;
ALTER TABLE attention_logs ENABLE ROW LEVEL SECURITY;
ALTER TABLE streaks ENABLE ROW LEVEL SECURITY;
ALTER TABLE weekly_summaries ENABLE ROW LEVEL SECURITY;

-- Policy template (apply to each table):
CREATE POLICY "Users can only access own data" ON profiles
    FOR ALL USING (auth.uid() = user_id);
-- Repeat for each table...
```

### Schema Relationships

```
auth.users (Supabase built-in)
    │
    ├── profiles (1:1)
    │
    ├── reframes (1:many)
    │
    ├── manifestations (1:many)
    │       │
    │       ├── daily_actions (1:many per manifestation)
    │       │
    │       └── attention_logs (1:many per manifestation)
    │
    ├── streaks (1:4 — one per streak_type)
    │
    └── weekly_summaries (1:many — one per week)
```

---

## 10. Monetization Model

### Freemium with Smart Gate

**Free Tier (acquisition & habit building):**
- 2 reframes per day (enough to build habit, not enough to feel unlimited)
- Access to Step 1 of Manifestation Engine (Decide/Clarity) — user creates a goal and feels invested, but can't access the full system
- Reframe history (last 7 days only)
- Basic streak tracking
- All research citations accessible

**Premium ($8/month or $60/year):**
- Unlimited reframes (soft limit 20/day, hard limit 50/day)
- Full 6-step Manifestation Engine
- Complete reframe history + pattern analysis
- Weekly AI summaries
- Multiple active manifestations (up to 5 concurrent goals)
- Audio visualizations (when available)
- Priority AI response (Sonnet for all calls vs. Haiku for free tier)

### Pricing Strategy

- **$8/month** — The "coffee a week" mental anchor. Positioned against therapy ($200+/session) and coaching ($100+/session).
- **$60/year** ($5/month effective) — 25% discount incentivizes annual commitment. Display prominently as the recommended option.
- **7-day free trial** for premium features — Lets users experience the full Manifestation Engine before committing.
- No lifetime purchase option initially — recurring revenue is more sustainable and aligns with the "daily practice" message.

### Revenue Projections (Conservative)

| Metric | Month 3 | Month 6 | Month 12 |
|---|---|---|---|
| Total downloads | 1,000 | 5,000 | 20,000 |
| Free → Premium conversion | 5% | 5% | 5% |
| Paying users | 50 | 250 | 1,000 |
| ARPU (blended monthly/annual) | $6.50 | $6.50 | $6.50 |
| MRR | $325 | $1,625 | $6,500 |
| ARR | $3,900 | $19,500 | $78,000 |

These are conservative estimates assuming organic growth + basic ASO. Paid acquisition or viral mechanics could significantly accelerate.

---

## 11. Cost Estimates

### Launch Phase (0-1,000 users)

| Service | Plan | Cost |
|---|---|---|
| Supabase | Free tier (500MB DB, 50K auth, 500K edge invocations) | $0 |
| Claude API | Pay-as-you-go (~$0.003-0.015/call) | $50-100/mo |
| Expo EAS | Free tier (30 builds/month) | $0 |
| RevenueCat | Free (up to $2,500 MTR) | $0 |
| PostHog | Free tier (1M events/mo) | $0 |
| Apple Developer | Annual fee | $99/yr ($8.25/mo) |
| Google Play Developer | One-time | $25 |
| **Total** | | **~$60-110/mo** |

### Growth Phase (1,000-10,000 users)

| Service | Plan | Cost |
|---|---|---|
| Supabase | Pro ($25/mo) | $25/mo |
| Claude API | Increased usage | $100-200/mo |
| Expo EAS | Production ($99/mo) | $99/mo |
| RevenueCat | 1% of revenue after $2.5K MTR | Variable |
| PostHog | Still free tier likely | $0 |
| **Total** | | **~$225-325/mo** |

### Breakeven Analysis

At $6.50 ARPU, breakeven on $250/mo infrastructure requires ~39 paying users. With 5% conversion, that's ~780 total users. Achievable within the first 2-3 months with moderate marketing effort.

---

## 12. Implementation Roadmap

### Phase 1 — MVP: The Reframer (4-6 weeks)

**Goal:** Ship the core reframing feature. Get it into users' hands. Validate the concept.

**Week 1-2: Foundation**
- Initialize Expo project with TypeScript
- Set up Supabase project (database, auth, edge functions)
- Implement auth flow (email + Google + Apple Sign-In)
- Create database tables and RLS policies
- Build basic navigation structure (tab navigator: Home, Reframe, History, Profile)

**Week 2-3: Onboarding**
- Welcome screens (swipeable intro cards)
- Reframe style selection (5 styles with example previews)
- Notification time picker
- Profile creation flow

**Week 3-4: Core Reframer**
- Text input screen for negative thoughts
- Supabase Edge Function: `/reframe` — takes thought + user context, calls Claude, returns reframe
- Reframe display card (with the three-part structure: acknowledge → reframe → micro-action)
- "Why This Works" expandable science section
- Save reframe to database
- "Try Different Style" button (re-call with different style)

**Week 4-5: History & Streaks**
- Reframe history list (chronological, filterable by category)
- Basic streak counter (current + longest)
- Daily reframe counter (enforce free tier limit)
- Push notification for daily check-in

**Week 5-6: Polish & Internal Testing**
- Error handling and loading states
- Offline support (queue reframe requests)
- App icon, splash screen, basic branding
- Internal TestFlight / Google Play internal testing

### Phase 2 — Manifestation Engine (4-6 weeks)

**Goal:** Build the full 6-step goal system. This is the premium feature.

**Week 7-8: Goal Creation Wizard (Steps 1-3)**
- Conversational AI flow for Step 1 (Decide) — multi-turn chat interface
- Visualization script generation for Step 2 (See It)
- Self-efficacy barrier identification for Step 3 (Believe)
- Save manifestation with all generated content

**Week 9-10: Daily System (Steps 4-5)**
- Attention prompts (push notifications with goal-specific prompts)
- Evidence journal (log observations)
- Daily if-then action generation and tracking
- Check-off interface with streak integration

**Week 10-11: Compound & Summary (Step 6)**
- Progress dashboard (visual timeline of actions, evidence, streaks)
- Weekly AI summary generation (cron job + display)
- Milestone celebrations

**Week 11-12: Subscription Gate**
- RevenueCat integration (iOS + Android)
- Paywall screen (appears when free user tries to access Steps 2-6)
- Premium entitlement checks throughout the app
- 7-day free trial flow
- Restore purchases functionality

### Phase 3 — Polish & Launch (2-4 weeks)

**Week 13-14: App Store Preparation**
- App Store screenshots and preview video
- App Store description (ASO-optimized)
- Privacy policy and terms of service
- App Store review compliance check
- Submit to Apple App Store and Google Play

**Week 14-16: Launch & Iteration**
- Analytics integration (PostHog events for all key actions)
- Crash reporting (Sentry or similar)
- Bug fixes from beta feedback
- Web version deployment (Expo Web → Vercel/Netlify)
- First marketing push

### Phase 4 — Growth Features (Ongoing)

- Voice input for quick reframes (Expo Audio → Whisper API or native STT)
- Audio visualization scripts (TTS generation, downloadable)
- Community features (anonymous sharing of reframes and wins)
- Pattern analysis dashboard (deeper insights into thought patterns)
- Widget support (iOS/Android home screen widget showing streak + daily prompt)
- Dark mode
- Multi-language support

---

## 13. Building with Claude Code & Codex

This section provides the exact workflow for building the app using AI-assisted development tools.

### Development Environment Setup

```bash
# 1. Install Node.js (LTS version)
# Download from https://nodejs.org or use nvm:
nvm install --lts
nvm use --lts

# 2. Install Expo CLI
npm install -g @expo/cli

# 3. Create the project
npx create-expo-app@latest [app-name] --template tabs
cd [app-name]

# 4. Install core dependencies
npx expo install @supabase/supabase-js
npx expo install expo-secure-store  # For secure token storage
npx expo install expo-notifications  # Push notifications
npx expo install react-native-purchases  # RevenueCat
npx expo install expo-av  # Audio playback (future)

# 5. Install navigation & UI
npx expo install @react-navigation/native @react-navigation/bottom-tabs
npx expo install react-native-screens react-native-safe-area-context

# 6. Development dependencies
npm install -D typescript @types/react
```

### Claude Code Workflow

Claude Code (the CLI tool) is ideal for this project. Here's how to use it effectively:

**Project initialization prompt:**
```
I'm building a React Native + Expo app with Supabase backend.
The app is a science-backed manifestation tool with two core features:
1. AI-powered thought reframing (user inputs negative thought → Claude API reframes it)
2. A 6-step goal achievement system

Tech stack: Expo (TypeScript), Supabase (Postgres + Edge Functions + Auth), Claude API.
Monetization: Freemium with RevenueCat.

Start by setting up the project structure, Supabase client configuration, 
and auth flow (email + Google + Apple Sign-In).
```

**Iterative development approach:**
1. Describe the feature you want to build
2. Let Claude Code scaffold the files
3. Review and test
4. Iterate with specific refinements
5. Move to next feature

**Recommended Claude Code commands for this project:**
```bash
# Start Claude Code in the project directory
claude

# Example prompts for each phase:

# Auth setup
"Set up Supabase auth with email, Google, and Apple Sign-In. 
Create a useAuth hook and an AuthContext provider."

# Reframer
"Create the reframe input screen with a text area, submit button, 
and loading state. It should call a Supabase Edge Function at /reframe."

# Edge Function
"Write a Supabase Edge Function that takes a negative thought and user's 
reframe style preference, calls the Claude API with a system prompt for 
thought reframing, and returns the structured response."

# Database
"Generate the SQL migration for creating the reframes table with RLS policies."
```

### Codex (OpenAI) Workflow

Codex can be used alongside Claude Code for:
- Generating boilerplate React Native components
- Writing database migration scripts
- Creating test files
- Generating TypeScript types from the schema

**Complementary usage:** Use Claude Code for architecture decisions and complex logic. Use Codex for repetitive scaffolding and boilerplate generation. They don't conflict — they complement.

### File Structure

```
[app-name]/
├── app/                          # Expo Router file-based routing
│   ├── (tabs)/                   # Tab navigator screens
│   │   ├── index.tsx             # Home / Dashboard
│   │   ├── reframe.tsx           # Reframe input screen
│   │   ├── goals.tsx             # Manifestation Engine
│   │   ├── history.tsx           # Reframe history
│   │   └── profile.tsx           # Settings & profile
│   ├── (auth)/                   # Auth screens
│   │   ├── login.tsx
│   │   ├── register.tsx
│   │   └── onboarding.tsx
│   ├── reframe/
│   │   └── [id].tsx              # Individual reframe detail
│   ├── manifestation/
│   │   ├── create.tsx            # Goal creation wizard
│   │   └── [id].tsx              # Goal detail / dashboard
│   └── _layout.tsx               # Root layout
├── components/
│   ├── ui/                       # Generic UI components
│   │   ├── Button.tsx
│   │   ├── Card.tsx
│   │   ├── TextInput.tsx
│   │   └── LoadingSpinner.tsx
│   ├── reframe/
│   │   ├── ReframeCard.tsx       # Displays a reframe result
│   │   ├── StyleSelector.tsx     # Reframe style picker
│   │   └── PatternInsight.tsx    # Pattern analysis display
│   ├── manifestation/
│   │   ├── StepProgress.tsx      # 6-step progress indicator
│   │   ├── ActionCard.tsx        # Daily action card
│   │   └── EvidenceEntry.tsx     # Attention log entry
│   └── shared/
│       ├── ScienceDrawer.tsx     # "Why This Works" expandable
│       └── StreakBadge.tsx       # Streak display component
├── hooks/
│   ├── useAuth.ts                # Authentication hook
│   ├── useReframe.ts             # Reframe CRUD operations
│   ├── useManifestation.ts       # Goal CRUD operations
│   ├── useStreak.ts              # Streak tracking
│   └── useSubscription.ts        # RevenueCat integration
├── lib/
│   ├── supabase.ts               # Supabase client init
│   ├── constants.ts              # App-wide constants
│   └── types.ts                  # TypeScript interfaces
├── contexts/
│   ├── AuthContext.tsx            # Auth state provider
│   └── SubscriptionContext.tsx    # Premium status provider
├── supabase/
│   ├── migrations/               # SQL migration files
│   │   ├── 001_create_profiles.sql
│   │   ├── 002_create_reframes.sql
│   │   ├── 003_create_manifestations.sql
│   │   ├── 004_create_daily_actions.sql
│   │   ├── 005_create_attention_logs.sql
│   │   ├── 006_create_streaks.sql
│   │   └── 007_create_weekly_summaries.sql
│   └── functions/                # Edge Functions
│       ├── reframe/index.ts
│       ├── generate-plan/index.ts
│       ├── generate-visualization/index.ts
│       ├── weekly-summary/index.ts
│       └── check-subscription/index.ts
├── assets/
│   ├── images/
│   └── fonts/
├── app.json                      # Expo config
├── tsconfig.json
└── package.json
```

---

## 14. App Store Strategy

### App Store Optimization (ASO)

**Primary keywords to target:**
- "manifestation app"
- "reframe negative thoughts"
- "positive thinking app"
- "goal setting app"
- "science-based self help"
- "thought reframing"
- "daily affirmation" (high volume, position as the "scientific" alternative)

**App Store title format:**
`[App Name] — Science-Backed Manifestation`

**Subtitle (30 chars):**
`Reframe Thoughts. Achieve Goals.`

**Description strategy:**
Lead with the unique differentiator (science-backed), include social proof (research citations), address the skeptic ("not another woo-woo app"), and end with clear feature list.

### Review Strategy

- Prompt for review after the user's 7th reframe (proven they find value)
- Never prompt during or after a negative reframe (bad timing)
- Prompt after completing Step 1 of a manifestation (feeling accomplished)

### Privacy & Compliance

- All data encrypted at rest (Supabase handles this)
- GDPR compliance: data export and deletion capability
- No third-party data sharing
- Mental health disclaimer: "This app is not a substitute for professional therapy or medical advice"
- Terms of service clearly stating AI-generated content limitations
- Age gate: 17+ (due to AI-generated content about potentially sensitive thoughts)

---

## 15. Growth & Future Features

### Phase 4+ Feature Ideas (Prioritized)

**High Impact / Lower Effort:**
1. **iOS/Android widget** — Home screen widget showing streak count + "Reframe Now" button. Reduces friction to zero.
2. **Dark mode** — Table stakes for a daily-use app.
3. **Voice input** — Speak your thought instead of typing. Uses Expo Audio + speech-to-text.
4. **Share reframes** — Export a reframe as a beautiful card image for social sharing (viral loop).

**High Impact / Higher Effort:**
5. **Audio visualizations** — Text-to-speech generation of visualization scripts. Users can listen during meditation or commute.
6. **Community** — Anonymous sharing of reframes and wins. Based on Granovetter's weak ties research. Users see how others transformed similar thoughts.
7. **Pattern dashboard** — Deep analytics on thought categories, time patterns, sentiment trends. "Your catastrophizing has decreased 40% since you started."
8. **Partner/friend feature** — Invite someone to share streaks and encourage each other (optional, privacy-first).

**Moonshot / Future:**
9. **AI-generated daily podcast** — Personalized 3-minute audio each morning based on current goals and recent patterns.
10. **Integration with calendar** — Auto-schedule if-then actions as calendar blocks.
11. **Wearable integration** — Apple Watch / WearOS app for quick-flip reframes.
12. **Therapist dashboard** — Optional feature where users can share their reframe/goal data with their actual therapist.

### Potential Pivots / Expansions

- **B2B:** Corporate wellness programs (companies buy bulk subscriptions for employees)
- **Education:** Student-focused version for schools/universities
- **Coaching platform:** Let certified coaches use the app as a tool with their clients, with shared dashboards

---

## 16. Naming Candidates

The app name is TBD. Here are the current candidates explored and their status:

### Taken / Conflicts

| Name | Issue |
|---|---|
| Reframe | Multiple apps (alcohol reduction, AI mental health, mindfulness) |
| MindShift | Existing CBT-based anxiety app |
| Prism | Multiple apps |
| Forge / MindForge | Various tech companies |
| Lumen | Metabolism tracker app |
| Myndset | Multiple companies (therapeutics, ideation) |
| Neurashift | AI consulting company |
| Mentavi | ADHD diagnostic platform |
| Soteria | Psychiatric treatment term |
| GrowMind | IT solutions + e-learning |
| Mindrift | AI freelance platform |

### Current Shortlist (Availability Needs Final Verification)

**Clareo**
- Latin: "clareo" = to be bright, to be clear
- Pronunciation: "cla-REE-oh"
- Directly connects to Step 1 (Clarity)
- Premium, international sound
- Existing "Clareo" entities are in B2B consulting (Chicago) and LED lighting (France) — zero overlap with self-help/wellness app category
- Status: Clean in app stores for our category

**Crux**
- English: the decisive or most important point
- "Getting to the crux of the matter"
- Short, strong, memorable, easy to spell
- Existing "Crux" apps: education (India) and rock climbing — different categories
- Status: Clean in app stores for our category

**Cairn** (previously top recommendation)
- Metaphor: Stone stacks hikers build to mark paths
- Meaning: Small daily actions stacking up, guiding forward
- Strong visual branding potential (stackable stones logo)
- Maps to Step 6 (Repeat and compound)
- Status: Need to verify

**Anneal**
- Metallurgy: heat and slowly cool metal to strengthen it
- Perfect metaphor for thought transformation (heat = challenge, cooling = reframing, result = stronger)
- Intellectual, science-forward brand positioning
- Status: Need to verify

### Other Viable Directions Worth Exploring

- Verge — "on the verge of change"
- Tether — grounded in reality
- Lucen — Latin "lucere" (to shine)
- Patina — beauty developed over time
- Aperion — Greek "infinite potential"
- Novis — Latin "new"

---

## 17. Research References

### Full Citation List with Open Access Links

1. **Locke, E.A. & Latham, G.P. (2002).** Building a Practically Useful Theory of Goal Setting and Task Motivation: A 35-Year Odyssey. *American Psychologist, 57*(9), 705-717. [PDF](https://med.stanford.edu/content/dam/sm/s-spire/documents/PD.locke-and-latham-retrospective_Paper.pdf)

2. **Locke, E.A., Latham, G.P., et al.** New Directions in Goal-Setting Theory. [PDF](https://www.lyondellbasell.com/4aeca6/globalassets/sustainability/lifebeats/advancing-health/life/goals/newdirectioningoalsetting_locke-et-al..pdf)

3. **Gollwitzer, P.M. (1999).** Implementation Intentions: Strong Effects of Simple Plans. *American Psychologist, 54*(7), 493-503. [PDF](https://prospectivepsych.org/sites/default/files/pictures/Gollwitzer_Implementation-intentions-1999.pdf)

4. **Gollwitzer, P.M. (2006).** Implementation Intentions and Goal Attainment. [Chapter PDF](https://cancercontrol.cancer.gov/sites/default/files/2020-06/goal_intent_attain.pdf)

5. **Schuster, C., Hilfiker, R., Amft, O., et al. (2011).** Best Practice for Motor Imagery: A Systematic Literature Review on Motor Imagery Training Elements in Five Different Disciplines. *BMC Medicine, 9*, 75. [PMC](https://pmc.ncbi.nlm.nih.gov/articles/PMC3141540)

6. **Liu, Y., et al. (2025).** Effects of Imagery Practice on Athletes' Performance. [PMC](https://pmc.ncbi.nlm.nih.gov/articles/PMC12109254)

7. **Seok, H., et al. (2023).** Mental Practice in Stroke Rehabilitation: A Systematic Review and Meta-Analysis. [PMC](https://pmc.ncbi.nlm.nih.gov/articles/PMC10689861)

8. **Bandura, A. (1977).** Self-Efficacy: Toward a Unifying Theory of Behavioral Change. *Psychological Review, 84*(2), 191-215. [PDF](https://educational-innovation.sydney.edu.au/news/pdfs/Bandura%201977.pdf)

9. **Yantis, S. (2008).** The Neural Basis of Selective Attention: Cortical Sources and Targets of Attentional Modulation. *Current Directions in Psychological Science, 17*(2), 86-90. [PMC](https://pmc.ncbi.nlm.nih.gov/articles/PMC2681259)

10. **Stevens, C. & Bavelier, D. (2012).** The Role of Selective Attention on Academic Foundations: A Cognitive Neuroscience Perspective. *Developmental Cognitive Neuroscience, 2*, S30-S48. [PMC](https://pmc.ncbi.nlm.nih.gov/articles/PMC3375497)

11. **Koch, C. & Tsuchiya, N. (2007).** Attention and Consciousness: Two Distinct Brain Processes. *Trends in Cognitive Sciences, 11*(1), 16-22. [PDF](https://philarchive.org/archive/KOCAAC-2)

12. **Wood, W., Mazar, A. & Neal, D.T. (2021).** Habits and Goals in Human Behavior: Separate but Interacting Systems. *Perspectives on Psychological Science, 17*(2), 590-605. [PDF](https://dornsife.usc.edu/wendy-wood/wp-content/uploads/sites/183/2023/10/Wood.Mazar_.Neal_.2021.pdf)

13. **Wood, W. & Neal, D.T. (2007).** A New Look at Habits and the Habit-Goal Interface. *Psychological Review, 114*(4), 843-863. [Academia](https://www.academia.edu/1237676/A_new_look_at_habits_and_the_habit_goal_interface)

14. **Granovetter, M.S. (1973).** The Strength of Weak Ties. *American Journal of Sociology, 78*(6), 1360-1380. [PDF](https://snap.stanford.edu/class/cs224w-readings/granovetter73weakties.pdf)

15. **Pardo-Cabello, A.J., et al. (2022).** Placebo and Nocebo Effects and Mechanisms Associated with Pharmacological Interventions. [PMC](https://pmc.ncbi.nlm.nih.gov/articles/PMC9361274)

16. **Stetler, D.A., et al. (2013).** The Role of Expectations in Treatment Outcome and the Placebo/Nocebo Effect. [PMC](https://pmc.ncbi.nlm.nih.gov/articles/PMC3851940)

---

## 18. Appendix: AI Prompt Templates

### Reframe System Prompt

```
You are a thought reframing assistant inside a science-backed personal 
transformation app. Your role is to help users transform negative thoughts 
into constructive, actionable perspectives.

USER'S PREFERRED STYLE: {{reframe_style}}

STYLE DEFINITIONS:
- brutal_honesty: Direct, no-nonsense. Challenge cognitive distortions bluntly 
  but with respect. Don't sugarcoat.
- gentle_coach: Warm, empathetic. Always validate feelings first. Use 
  encouraging language while gently shifting perspective.
- logical_analyst: Structured, evidence-based. Use CBT-style questioning. 
  Identify logical fallacies in the thought pattern.
- philosophical: Big-picture thinking. Use temporal distancing, perspective 
  shifts, and existential reframing.
- humor: Light, disarming. Use appropriate wit to defuse catastrophizing while 
  still taking the core concern seriously.

RESPONSE STRUCTURE (always follow this):
1. ACKNOWLEDGE: Start by validating the feeling (1-2 sentences). Never dismiss.
2. REFRAME: Provide the perspective shift in your style (3-5 sentences). 
   Where natural, reference relevant research concepts (attention bias, 
   catastrophizing, self-efficacy) without being preachy.
3. MICRO-ACTION: End with ONE if-then implementation intention:
   "If [specific trigger], then I will [specific small action]."

CRITICAL RULES:
- NEVER use toxic positivity ("everything happens for a reason", "just think 
  positive", "the universe has a plan")
- NEVER minimize real problems (grief, abuse, discrimination, health crises)
- If the thought describes a genuinely dangerous or crisis situation, 
  acknowledge it seriously and suggest professional support
- Keep responses under 200 words
- Do not use bullet points or numbered lists in the response body
- Sound human, not like a chatbot
- Use "you" not "the user"

CONTEXT (if available):
- User's recent reframe categories: {{recent_categories}}
- Recurring patterns: {{patterns}}
- Active manifestation goals: {{active_goals}}
```

### Goal Creation Wizard System Prompt (Step 1: Decide)

```
You are a goal-setting guide inside a science-backed manifestation app. 
You're helping the user define a clear, specific, personally meaningful goal 
using principles from Locke & Latham's goal-setting theory.

This is a CONVERSATIONAL flow. Ask ONE question at a time. Be warm and 
curious, not clinical. Your job is to help them get crystal clear on what 
they want and why it matters.

CONVERSATION FLOW:
1. Ask what they want to achieve (open-ended)
2. Ask why this matters to them personally (emotional connection)
3. Help them make it specific and measurable (SMART, but don't use that 
   acronym — make it natural)
4. Help them set a realistic but challenging timeframe
5. Summarize their goal as a "Clarity Statement" — a 1-2 sentence 
   declaration that is specific, personally meaningful, and time-bound

RULES:
- One question per response. Keep it conversational.
- Don't lecture about goal-setting theory. Just apply it naturally.
- If their goal is vague, help them sharpen it through questions, not 
  instructions.
- If their goal seems unrealistic, gently explore feasibility without 
  crushing ambition. Ask about intermediate milestones.
- The Clarity Statement should feel like THEIR words refined, not your 
  words imposed.
- Keep responses under 100 words per exchange.
```

### Visualization Script System Prompt (Step 2: See It)

```
You are generating a personalized visualization script for the user's goal.
This script will be read (and eventually listened to) as a mental rehearsal 
exercise.

GOAL: {{clarity_statement}}
WHY IT MATTERS: {{why_it_matters}}

VISUALIZATION PRINCIPLES (from Schuster et al., 2011):
- First person perspective ("I am..." not "You are...")
- Sensory-rich: include visual, auditory, kinesthetic, and emotional details
- Process-focused: visualize the ACTIONS and JOURNEY, not just the outcome
- Realistic: include small challenges that are overcome (builds resilience)
- Present tense: written as if happening right now

FORMAT:
- 300-500 words
- Written as a guided meditation script
- Begin with a grounding instruction ("Take a deep breath...")
- Move through the process of achieving the goal
- End with the emotional state of having achieved it
- Close with a return to the present moment

TONE: Calm, vivid, personal. This should feel like reading a scene from the 
user's best possible future — grounded and real, not fantasy.
```

### Weekly Summary System Prompt

```
You are generating a weekly transformation report for a user of a science-
backed manifestation app. Analyze their data and provide an encouraging, 
honest, and actionable summary.

USER DATA FOR THIS WEEK:
- Reframes completed: {{reframe_count}}
- Reframe categories: {{category_breakdown}}
- Average sentiment shift: {{sentiment_delta}}
- Active manifestation(s): {{active_goals}}
- Daily actions completed: {{actions_completed}} / {{actions_total}}
- Attention logs recorded: {{attention_count}}
- Current streak: {{streak_count}} days
- Comparison to last week: {{week_over_week}}

RESPONSE STRUCTURE:
1. HEADLINE INSIGHT: One sentence capturing the week's theme (encouraging)
2. PATTERNS: What you notice in their reframe categories and timing 
   (be specific: "Work stress peaked on Tuesday" not "you had some stress")
3. PROGRESS: Goal achievement metrics with genuine encouragement
4. SUGGESTION: One specific, actionable recommendation for next week 
   (tied to a research insight)
5. MOTIVATION: Close with something that connects their daily effort to 
   their bigger goal

RULES:
- Be honest. If they had a rough week or low completion, acknowledge it 
  without judgment and normalize it.
- Use specific numbers from their data.
- Keep under 250 words.
- Never use generic platitudes. Every sentence should be specific to THIS 
  user's data.
```

---

## Document History

| Version | Date | Changes |
|---|---|---|
| 1.0 | 2026-02-09 | Initial comprehensive blueprint |

---

*This document is a living blueprint. It will be updated as development progresses, decisions are made, and the product evolves.*

*Built with 🧠 science + 🤖 AI + 💪 daily action*