b2c-app-builder 0.221.39

package/catalog/generated/firstparty/knowledge/experience/experience-cards/intent-mirroring-card.md

@@ -0,0 +1,36 @@
+# Intent Mirroring Card
+
+Part of the Experience Card deck. Read [`../experience-cards.md`](../experience-cards.md) first — it carries the card shape, the summary table, the Ethics Ladder/attestation contract, and the per-card routing. Apply cards via `EMOTIONAL_DESIGN.md` per `emotional-design-system.md`; HIGH-risk mechanics carry the full ethics contract in `ethics-guardrail.md`.
+
+## Intent Mirroring Card
+
+**One-liner.** A deliberate pause that reflects the user's declared goal back in their own
+words — before effort or payment — turning a transactional interaction into a moment of
+self-recognition that deepens motivation and perceived fit.
+
+**Risk tier.** MEDIUM — canonical in the [`../experience-cards.md`](../experience-cards.md) routing table and [`../ethics-guardrail.md`](../ethics-guardrail.md) §3.
+
+### Bright Line / Dark Line
+
+**Bright line.** The mirror reflects the user's own stated words or explicit selections — their
+actual goal, obstacle, or declared intent — back to them, at a quiet moment, before asking for
+effort or payment, in service of making them more likely to achieve what they came to the app
+to achieve. The user feels recognized, not analyzed.
+
+**Dark line.** Mirroring an inferred, escalated, or manufactured emotional state the user did
+not explicitly express — to manufacture anxiety, urgency, or guilt that primes a purchase.
+Using the mirror on a cancellation or downgrade path as a confirmshaming device. Both are the
+mirror mechanism weaponized as a retention dark pattern.
+
+### Full Card — Served by the Retention Mechanics MCP
+
+This bundled file is the frozen free-layer stub: the deck's content boundary is
+explicit as of Phase 3 (see `../experience-cards.md` §Live deck access). The full
+annotation — citation-graded psychology, mechanism steps, real-app examples,
+producer recipe, auditor signals, measurement events, mobile implementation +
+reduced-motion recipes, and the guardrail's on-device verification contract — is
+served live: `retention_get_mechanic("intent-mirroring")`, then
+`retention_get_ethics_ladder("intent-mirroring")` before building. Without the server,
+apply the deck contract in [`../experience-cards.md`](../experience-cards.md) and
+the ethics contract in [`../ethics-guardrail.md`](../ethics-guardrail.md); the
+Ethics Ladder's attestation requirements apply unchanged at this card's tier.

package/catalog/generated/firstparty/knowledge/experience/experience-cards/mastery-and-status-card.md

@@ -0,0 +1,43 @@
+# Mastery and Status Card
+
+Part of the Experience Card deck. Read [`../experience-cards.md`](../experience-cards.md) first — it carries the card shape, the summary table, the Ethics Ladder/attestation contract, and the per-card routing. Apply cards via `EMOTIONAL_DESIGN.md` per `emotional-design-system.md`; HIGH-risk mechanics carry the full ethics contract in `ethics-guardrail.md`.
+
+## Mastery and Status Card
+
+**One-liner.** Earned status through visible skill levels, domain-vocabulary badges, and
+social proof of progress gives users an intrinsic and social reason to continue beyond the
+functional goal — because earned identity is self-reinforcing and publicly displayable.
+
+**Risk tier.** MEDIUM — canonical in the [`../experience-cards.md`](../experience-cards.md) routing table and [`../ethics-guardrail.md`](../ethics-guardrail.md) §3.
+
+### Bright Line / Dark Line
+
+**Bright line.** The level label reflects genuine skill improvement measured by a real
+behavioral metric documented in `analytics/ANALYTICS.md` and `engineering/TECH_SPEC.md`; the vocabulary is drawn
+from the user's skill domain and feels earned; progress is always visible; and social display
+is optional, with a clear opt-out.
+
+**Dark line.** The mastery metric is a proxy engagement score with no relationship to real
+skill improvement; level-up reveals gated behind a subscription paywall; near-level progress
+bars hidden or reset on cancellation; the level-up reveal fires on the same screen as a
+paywall CTA, weaponizing earned pride as a purchase trigger.
+
+### Full Card — Served by the Retention Mechanics MCP
+
+This bundled file is the frozen free-layer stub: the deck's content boundary is
+explicit as of Phase 3 (see `../experience-cards.md` §Live deck access). The full
+annotation — citation-graded psychology, mechanism steps, real-app examples,
+producer recipe, auditor signals, measurement events, mobile implementation +
+reduced-motion recipes, and the guardrail's on-device verification contract — is
+served live: `retention_get_mechanic("mastery-and-status")`, then
+`retention_get_ethics_ladder("mastery-and-status")` before building. Without the server,
+apply the deck contract in [`../experience-cards.md`](../experience-cards.md) and
+the ethics contract in [`../ethics-guardrail.md`](../ethics-guardrail.md); the
+Ethics Ladder's attestation requirements apply unchanged at this card's tier.
+
+### Motion Binding (canon)
+
+The celebrate-family spring canon stays pinned here so `check:motion-contract`
+keeps this value undriftable: `.spring(response: 0.5, dampingFraction: 0.7)`. The full
+platform recipes (SwiftUI, Flutter, React Native, web) and their reduced-motion
+fallbacks are in the MCP-served card.

package/catalog/generated/firstparty/knowledge/experience/experience-cards/peak-end-card.md

@@ -0,0 +1,44 @@
+# Peak-End Card
+
+Part of the Experience Card deck. Read [`../experience-cards.md`](../experience-cards.md) first — it carries the card shape, the summary table, the Ethics Ladder/attestation contract, and the per-card routing. Apply cards via `EMOTIONAL_DESIGN.md` per `emotional-design-system.md`; HIGH-risk mechanics carry the full ethics contract in `ethics-guardrail.md`.
+
+## Peak-End Card
+
+**One-liner.** Engineer one unforgettable in-session peak and a strong close; the remembering
+self makes every re-engagement decision based on those two moments, not the average.
+
+**Risk tier.** MEDIUM — canonical in the [`../experience-cards.md`](../experience-cards.md) routing table and [`../ethics-guardrail.md`](../ethics-guardrail.md) §3.
+
+### Bright Line / Dark Line
+
+**Bright line.** The peak moment celebrates the user's actual earned result — a computation,
+milestone, or insight grounded in their real actions and data — and the session close confirms
+their genuine accomplishment. Every personalization field referenced in peak or close copy must
+trace to a field the user explicitly provided or a result the product computed from their real
+activity.
+
+**Dark line.** Manufacturing an emotional peak from fabricated or inflated results ("You're in
+the top 5% of users!" on day one with no real comparison basis); using the peak as a warm-up
+for an immediate hard-sell paywall on the same screen; triggering the peak animation on every
+tap regardless of earned result; using the close state to inject anxiety ("Don't lose your
+progress — subscribe now") rather than completing the session with satisfaction.
+
+### Full Card — Served by the Retention Mechanics MCP
+
+This bundled file is the frozen free-layer stub: the deck's content boundary is
+explicit as of Phase 3 (see `../experience-cards.md` §Live deck access). The full
+annotation — citation-graded psychology, mechanism steps, real-app examples,
+producer recipe, auditor signals, measurement events, mobile implementation +
+reduced-motion recipes, and the guardrail's on-device verification contract — is
+served live: `retention_get_mechanic("peak-end")`, then
+`retention_get_ethics_ladder("peak-end")` before building. Without the server,
+apply the deck contract in [`../experience-cards.md`](../experience-cards.md) and
+the ethics contract in [`../ethics-guardrail.md`](../ethics-guardrail.md); the
+Ethics Ladder's attestation requirements apply unchanged at this card's tier.
+
+### Motion Binding (canon)
+
+The celebrate-family spring canon stays pinned here so `check:motion-contract`
+keeps this value undriftable: `.spring(response: 0.45, dampingFraction: 0.7)`. The full
+platform recipes (SwiftUI, Flutter, React Native, web) and their reduced-motion
+fallbacks are in the MCP-served card.

package/catalog/generated/firstparty/knowledge/experience/experience-cards/perceived-effort-delay-card.md

@@ -0,0 +1,37 @@
+# Perceived Effort Delay Card
+
+Part of the Experience Card deck. Read [`../experience-cards.md`](../experience-cards.md) first — it carries the card shape, the summary table, the Ethics Ladder/attestation contract, and the per-card routing. Apply cards via `EMOTIONAL_DESIGN.md` per `emotional-design-system.md`; HIGH-risk mechanics carry the full ethics contract in `ethics-guardrail.md`.
+
+## Perceived Effort Delay Card
+
+**One-liner.** Show the system working through the user's specific inputs — a narrated
+processing moment that converts wait time into perceived craftsmanship and raises willingness
+to pay.
+
+**Risk tier.** MEDIUM — canonical in the [`../experience-cards.md`](../experience-cards.md) routing table and [`../ethics-guardrail.md`](../ethics-guardrail.md) §3.
+
+### Bright Line / Dark Line
+
+**Bright line.** The processing delay shows the user evidence of real work — a real API call,
+real data processing, or real UI composition — narrated in the user's own language. The user
+can cancel at any time. Accuracy of the steps is upheld: if the system says "Analyzing your 5
+goals," it used all 5 goals in the computation.
+
+**Dark line.** An artificial delay added by a sleep timer or arbitrary `setTimeout` with no
+underlying computation — pure theater. Equally: a progress bar that advances randomly
+regardless of real computation stages (a fake progress bar). A discovered fake delay produces
+a trust collapse worse than no delay at all, triggers "just a spinner" and "doesn't actually
+analyze" review clusters, and is a compliance veto.
+
+### Full Card — Served by the Retention Mechanics MCP
+
+This bundled file is the frozen free-layer stub: the deck's content boundary is
+explicit as of Phase 3 (see `../experience-cards.md` §Live deck access). The full
+annotation — citation-graded psychology, mechanism steps, real-app examples,
+producer recipe, auditor signals, measurement events, mobile implementation +
+reduced-motion recipes, and the guardrail's on-device verification contract — is
+served live: `retention_get_mechanic("perceived-effort-delay")`, then
+`retention_get_ethics_ladder("perceived-effort-delay")` before building. Without the server,
+apply the deck contract in [`../experience-cards.md`](../experience-cards.md) and
+the ethics contract in [`../ethics-guardrail.md`](../ethics-guardrail.md); the
+Ethics Ladder's attestation requirements apply unchanged at this card's tier.

package/catalog/generated/firstparty/knowledge/experience/experience-cards/reciprocity-card.md

@@ -0,0 +1,38 @@
+# Reciprocity Card
+
+Part of the Experience Card deck. Read [`../experience-cards.md`](../experience-cards.md) first — it carries the card shape, the summary table, the Ethics Ladder/attestation contract, and the per-card routing. Apply cards via `EMOTIONAL_DESIGN.md` per `emotional-design-system.md`; HIGH-risk mechanics carry the full ethics contract in `ethics-guardrail.md`.
+
+## Reciprocity Card
+
+**One-liner.** Deliver unexpected value before asking for anything in return — a free
+capability, a personalised insight, a gift — so the user feels genuinely obligated to give
+back: an upgrade, a permission grant, a share.
+
+**Risk tier.** MEDIUM — canonical in the [`../experience-cards.md`](../experience-cards.md) routing table and [`../ethics-guardrail.md`](../ethics-guardrail.md) §3.
+
+### Bright Line / Dark Line
+
+**Bright line.** The gift is a complete, self-contained, genuinely useful capability or insight
+derived from the user's own inputs — delivered without announcement, before any ask — so the
+user's obligation flows naturally from gratitude. The ask that follows is proportional
+(permission, soft upgrade, or share) and occurs at least 2 seconds after the gift lands. The
+user can dismiss the ask without losing the gift they already received.
+
+**Dark line.** The Reciprocity Card crosses into manipulation when: (1) the "gift" withholds
+the genuinely useful part behind a paywall — the user receives a teaser, not a complete value;
+(2) the ask fires on a distressed screen; (3) the gift is deployed on the cancel or unsubscribe
+flow as a retention tactic; (4) the gift is announced as a marketing tactic rather than
+delivered as a surprise.
+
+### Full Card — Served by the Retention Mechanics MCP
+
+This bundled file is the frozen free-layer stub: the deck's content boundary is
+explicit as of Phase 3 (see `../experience-cards.md` §Live deck access). The full
+annotation — citation-graded psychology, mechanism steps, real-app examples,
+producer recipe, auditor signals, measurement events, mobile implementation +
+reduced-motion recipes, and the guardrail's on-device verification contract — is
+served live: `retention_get_mechanic("reciprocity")`, then
+`retention_get_ethics_ladder("reciprocity")` before building. Without the server,
+apply the deck contract in [`../experience-cards.md`](../experience-cards.md) and
+the ethics contract in [`../ethics-guardrail.md`](../ethics-guardrail.md); the
+Ethics Ladder's attestation requirements apply unchanged at this card's tier.

package/catalog/generated/firstparty/knowledge/experience/experience-cards/recovery-and-trust-repair-card.md

@@ -0,0 +1,40 @@
+# Recovery and Trust Repair Card
+
+Part of the Experience Card deck. Read [`../experience-cards.md`](../experience-cards.md) first — it carries the card shape, the summary table, the Ethics Ladder/attestation contract, and the per-card routing. Apply cards via `EMOTIONAL_DESIGN.md` per `emotional-design-system.md`; HIGH-risk mechanics carry the full ethics contract in `ethics-guardrail.md`.
+
+## Recovery and Trust Repair Card
+
+**One-liner.** A product that handles errors, lapses, failed payments, and crashes with speed,
+transparency, and a restorative offer builds more durable trust than one that simply never
+fails, because users remember the recovery more vividly than the failure itself.
+
+**Risk tier.** MEDIUM — canonical in the [`../experience-cards.md`](../experience-cards.md) routing table and [`../ethics-guardrail.md`](../ethics-guardrail.md) §3.
+
+### Bright Line / Dark Line
+
+**Bright line.** Use this card to restore the user's path toward their stated goal as quickly
+and honestly as possible. A recovery that names what happened, restores what can be restored,
+and offers an unconditional gesture when the failure was the product's fault serves the user's
+real interest. The peak-end rule ensures the memory of the recovery — not the failure —
+becomes the dominant emotional impression, which aligns user benefit with product retention.
+
+**Dark line.** Exploiting the user's distress at a failure moment to drive a spend action:
+surfacing a paywall, an IAP offer, or a streak-restore purchase inside or immediately after
+the failure/grief screen. This is the recovery equivalent of the prohibited pattern in
+`ethics-guardrail.md` (spend prompt inside a grief screen). It is a non-negotiable prohibited
+pattern under Apple App Store Review Guidelines §4.0 (exploit emotions) and FTC dark-patterns
+enforcement. A second dark-pattern variant: using operational transparency theatrics during a
+fake repair — showing animated "fixing your account" steps when no actual repair is occurring.
+
+### Full Card — Served by the Retention Mechanics MCP
+
+This bundled file is the frozen free-layer stub: the deck's content boundary is
+explicit as of Phase 3 (see `../experience-cards.md` §Live deck access). The full
+annotation — citation-graded psychology, mechanism steps, real-app examples,
+producer recipe, auditor signals, measurement events, mobile implementation +
+reduced-motion recipes, and the guardrail's on-device verification contract — is
+served live: `retention_get_mechanic("recovery-and-trust-repair")`, then
+`retention_get_ethics_ladder("recovery-and-trust-repair")` before building. Without the server,
+apply the deck contract in [`../experience-cards.md`](../experience-cards.md) and
+the ethics contract in [`../ethics-guardrail.md`](../ethics-guardrail.md); the
+Ethics Ladder's attestation requirements apply unchanged at this card's tier.

package/catalog/generated/firstparty/knowledge/experience/experience-cards/streak-and-loss-aversion-card.md

@@ -0,0 +1,37 @@
+# Streak and Loss Aversion Card
+
+Part of the Experience Card deck. Read [`../experience-cards.md`](../experience-cards.md) first — it carries the card shape, the summary table, the Ethics Ladder/attestation contract, and the per-card routing. Apply cards via `EMOTIONAL_DESIGN.md` per `emotional-design-system.md`; HIGH-risk mechanics carry the full ethics contract in `ethics-guardrail.md`.
+
+## Streak and Loss Aversion Card
+
+**One-liner.** Turn an earned run of days into a possession the user refuses to surrender —
+driving daily return visits through the asymmetric pain of losing something already owned.
+
+**Risk tier.** HIGH — canonical in the [`../experience-cards.md`](../experience-cards.md) routing table and [`../ethics-guardrail.md`](../ethics-guardrail.md) §3.
+
+### Bright Line / Dark Line
+
+**Bright line.** The streak is anchored to the user's stated goal (captured by the Commitment
+Card) and measures completion of a core action that delivers real value toward that goal.
+Forgiveness mechanics are freely accessible — at minimum a grace period, without requiring
+payment. Notification copy names the goal, not just the streak length. The streak serves the
+user's retention of their habit, not the app's engagement metrics.
+
+**Dark line.** The streak is anchored to any app open or passive action rather than the user's
+goal-relevant core behavior. Forgiveness is exclusively paid. A repair purchase is surfaced
+immediately after a break with no grace period. Notification copy uses guilt, countdown panic,
+or manufactured shame ("You're failing your goal — don't break now"). In a wellness app, streak
+mechanics produce the anxiety the product is sold to reduce.
+
+### Full Card — Served by the Retention Mechanics MCP
+
+This bundled file is the frozen free-layer stub: the deck's content boundary is
+explicit as of Phase 3 (see `../experience-cards.md` §Live deck access). The full
+annotation — citation-graded psychology, mechanism steps, real-app examples,
+producer recipe, auditor signals, measurement events, mobile implementation +
+reduced-motion recipes, and the guardrail's on-device verification contract — is
+served live: `retention_get_mechanic("streak-and-loss-aversion")`, then
+`retention_get_ethics_ladder("streak-and-loss-aversion")` before building. Without the server,
+apply the deck contract in [`../experience-cards.md`](../experience-cards.md) and
+the ethics contract in [`../ethics-guardrail.md`](../ethics-guardrail.md); the
+Ethics Ladder's attestation requirements apply unchanged at this card's tier.

package/catalog/generated/firstparty/knowledge/experience/experience-cards/variable-reward-card.md

@@ -0,0 +1,42 @@
+# Variable Reward Card
+
+Part of the Experience Card deck. Read [`../experience-cards.md`](../experience-cards.md) first — it carries the card shape, the summary table, the Ethics Ladder/attestation contract, and the per-card routing. Apply cards via `EMOTIONAL_DESIGN.md` per `emotional-design-system.md`; HIGH-risk mechanics carry the full ethics contract in `ethics-guardrail.md`.
+
+## Variable Reward Card
+
+**One-liner.** Unpredictable positive outcomes delivered after an anticipation window produce
+the highest engagement rate of any reinforcement schedule — because dopamine fires on the
+wait, not the receipt.
+
+**Risk tier.** HIGH — canonical in the [`../experience-cards.md`](../experience-cards.md) routing table and [`../ethics-guardrail.md`](../ethics-guardrail.md) §3.
+
+### Bright Line / Dark Line
+
+**Bright line.** The variation is real and every possible result serves the user's stated goal.
+The user could receive any result in the range and feel the app worked for them. The mechanic
+keeps users curious about their own progress, not anxious about an outcome they paid for.
+
+**Dark line.** Applying variable reward to paid spend (loot boxes); engineering near-miss
+signals to increase retry behavior; coupling the dopamine anticipation window to a spend prompt
+on the same screen; using cosmetic variation to manufacture engagement without genuine
+personalization.
+
+### Full Card — Served by the Retention Mechanics MCP
+
+This bundled file is the frozen free-layer stub: the deck's content boundary is
+explicit as of Phase 3 (see `../experience-cards.md` §Live deck access). The full
+annotation — citation-graded psychology, mechanism steps, real-app examples,
+producer recipe, auditor signals, measurement events, mobile implementation +
+reduced-motion recipes, and the guardrail's on-device verification contract — is
+served live: `retention_get_mechanic("variable-reward")`, then
+`retention_get_ethics_ladder("variable-reward")` before building. Without the server,
+apply the deck contract in [`../experience-cards.md`](../experience-cards.md) and
+the ethics contract in [`../ethics-guardrail.md`](../ethics-guardrail.md); the
+Ethics Ladder's attestation requirements apply unchanged at this card's tier.
+
+### Motion Binding (canon)
+
+The celebrate-family spring canon stays pinned here so `check:motion-contract`
+keeps this value undriftable: `.spring(response: 0.5, dampingFraction: 0.6)`. The full
+platform recipes (SwiftUI, Flutter, React Native, web) and their reduced-motion
+fallbacks are in the MCP-served card.

package/catalog/generated/firstparty/knowledge/experience/experience-cards.md

@@ -0,0 +1,172 @@
+# Experience Cards
+
+Use this when designing, auditing, or implementing emotionally charged moments in a B2C mobile
+app. These cards are the canonical deck for the B2C App Builder skill. Each card
+names a psychological mechanism, draws a bright/dark line, and gives you a deterministic
+checklist for audit and ship.
+## Applicability And Evidence Boundary
+
+Card selection is conditional; the discipline of making and reviewing that selection is core. Calm, predictable feedback and immediate completion can be the right emotional experience. Evaluate the actual user need before selecting novelty, anticipation, a commitment, or a reveal. Do not manufacture a behavioral loop to fill a card checklist.
+
+Psychological theories and examples below motivate hypotheses. They do not establish a universal effect size, a guaranteed star-level improvement, or a measured benefit in this product. Keep observed user response separate from intended tone. Every applied card retains its ethics, truthful progress, accessibility and measurement obligations; marking another card non-applicable waives none of them.
+
+
+**This file is the index.** Each file under [`experience-cards/`](./experience-cards/commitment-card.md)
+is a compact routing stub — title, one-liner, risk tier, bright/dark lines, and a
+served-by-MCP pointer (the three motion-bearing cards also pin their celebrate-family spring
+canon for `check:motion-contract`). The full per-card specs (psychology + canonical research,
+mechanism steps, real app examples, producer recipe, auditor signals, measurement events,
+mobile implementation + reduced-motion, pairings) are served by the Retention Mechanics MCP —
+see Live deck access below. Load **only the cards in scope for the current moment**; the
+stubs are cheap, and the MCP serves full depth per card so loading the whole deck for a
+single paywall pause is never necessary.
+
+**Live deck access (Retention Mechanics MCP).** The deck behind this index is queryable live
+when a `retention-mechanics` MCP server is connected (first-party; tools named
+`retention_*`; registered as `retention-mechanics-mcp`). When it is connected, prefer it for
+card routing and per-card content — it serves the maintained, versioned deck (per-card
+`card_version`/`last_reviewed`, canonicalized risk tiers, attestation scaffolds) and exceeds
+the bundled twelve in both count and depth: the bundled files route, the MCP serves the full
+annotation for every card:
+
+- `retention_search_mechanics` — search the deck by product moment, with funnel-stage and
+  ethics-risk-ceiling filters; every result carries the card's do-not-use conditions, so a
+  vocabulary match cannot pass as a situation-fit diagnosis.
+- `retention_get_mechanic` — one full card, citation confidence passed through verbatim and
+  real-app examples marked `illustrative_unverified`.
+- `retention_get_ethics_ladder` — the risk tier, bright/dark lines, the attestation fields
+  the validator requires at that tier, and a scaffold whose narrative fields are
+  intentionally empty (the applying team writes them for its own surface).
+
+The bundled files under `experience-cards/` are the frozen free-layer stubs of the deck's
+twelve foundational cards and remain the offline contract at stub depth: every validator and
+`npm run audit:ci` run reads the bundled files, never the network, and when the server is not
+connected this index and the stubs stay authoritative for routing, risk tiers, the bright/dark
+ship-gates, and the pinned motion canon. The full per-card spec is authoritative only through
+a live MCP connection — the content boundary between the free layer and the maintained deck
+is explicit as of Phase 3 of the deck migration.
+
+**Pre-requisites.** Load these references before applying any card; do not duplicate their
+content here:
+
+- `knowledge/experience/eleven-star-experience.md` — star-ladder; every card maps to a level
+- `knowledge/design/quality-lens.md` — emotional job, not a generic SaaS wrapper
+- `knowledge/experience/onboarding-conversion.md` — paywall timing, App Review popup, consent
+- `knowledge/data/analytics-attribution.md` — every emotional moment needs a named PostHog event
+- `knowledge/design/design-room.md` / `knowledge/design/design-visual-system.md` — motion tokens and
+  reduced-motion rules; every delight animation needs an OS-level fallback
+- `knowledge/process/failure-cards.md` — dark-pattern violations become failure cards
+- `knowledge/experience/ethics-guardrail.md` — Guardrail Contract, regulatory landscape, risk table
+
+---
+
+When no card applies, record the following fenced YAML inside the existing `EMOTIONAL_DESIGN.md`. Each text field must explain the actual product decision (at least 30 characters, no placeholders). The audit independently records its decision in `EMOTIONAL_AUDIT.md`; an empty map alone is insufficient. This selects zero cards, not a skipped ethics review. A selection cannot coexist with applied card blocks.
+
+```yaml
+experience_card_selection:
+  status: not_applicable
+  user_job: Locate the household document and read its renewal date.
+  rationale: Predictable retrieval needs no commitment, reward, wait or intent mirror.
+  alternative: Show the requested record immediately with clear status and recovery.
+```
+
+
+## Card Routing
+
+One line per card: when to load it, and where the full spec lives.
+
+| Card                       | Load when                                                                                        | Risk     | Spec                                                                                                               |
+| -------------------------- | ------------------------------------------------------------------------------------------------ | -------- | ------------------------------------------------------------------------------------------------------------------ |
+| Commitment                 | A voluntary user-authored goal can anchor the journey (onboarding goal question, plan setup)     | MEDIUM   | [`experience-cards/commitment-card.md`](./experience-cards/commitment-card.md)                                     |
+| Variable Reward            | An outcome genuinely varies and a reveal moment exists (results, generations, daily pulls)       | **HIGH** | [`experience-cards/variable-reward-card.md`](./experience-cards/variable-reward-card.md)                           |
+| Perceived Effort Delay     | Real computation runs on the user's behalf and can be shown honestly (plan generation, analysis) | MEDIUM   | [`experience-cards/perceived-effort-delay-card.md`](./experience-cards/perceived-effort-delay-card.md)             |
+| Intent Mirroring           | A pause can reflect the user's own stated goal back (pre-paywall, return sessions)               | MEDIUM   | [`experience-cards/intent-mirroring-card.md`](./experience-cards/intent-mirroring-card.md)                         |
+| Endowed Progress           | Real prior progress exists to surface before a multi-step task (setup, levels, profiles)         | LOW      | [`experience-cards/endowed-progress-card.md`](./experience-cards/endowed-progress-card.md)                         |
+| Peak-End                   | A session has a natural emotional peak and a designable ending (completions, summaries)          | MEDIUM   | [`experience-cards/peak-end-card.md`](./experience-cards/peak-end-card.md)                                         |
+| Streak & Loss Aversion     | Continuity itself is the value and a free recovery path will ship (habits, daily practice)       | **HIGH** | [`experience-cards/streak-and-loss-aversion-card.md`](./experience-cards/streak-and-loss-aversion-card.md)         |
+| Reciprocity                | An unprompted, real gift can precede any ask (bonus content, surprise upgrades)                  | MEDIUM   | [`experience-cards/reciprocity-card.md`](./experience-cards/reciprocity-card.md)                                   |
+| Identity & Self-Expression | Users can shape something that reflects who they are (avatars, spaces, collections)              | MEDIUM   | [`experience-cards/identity-and-self-expression-card.md`](./experience-cards/identity-and-self-expression-card.md) |
+| Fresh Start                | A temporal landmark can reopen a lapsed journey without guilt (new week/month, post-milestone)   | MEDIUM   | [`experience-cards/fresh-start-card.md`](./experience-cards/fresh-start-card.md)                                   |
+| Mastery & Status           | Skill genuinely grows and earned tiers can be shown truthfully (progressions, badges)            | MEDIUM   | [`experience-cards/mastery-and-status-card.md`](./experience-cards/mastery-and-status-card.md)                     |
+| Recovery & Trust Repair    | A failure state (error, crash, failed payment, lapsed streak) needs to win trust back            | MEDIUM   | [`experience-cards/recovery-and-trust-repair-card.md`](./experience-cards/recovery-and-trust-repair-card.md)       |
+
+---
+
+## Card Shape
+
+**Bundled stub field set** — what each file under `experience-cards/` carries on disk:
+
+| Field                       | Meaning                                                                                                                        |
+| --------------------------- | ------------------------------------------------------------------------------------------------------------------------------ |
+| **One-liner**               | Single-sentence description of what the card does                                                                              |
+| **Risk tier**               | LOW/MEDIUM/HIGH, mirroring this index's routing table and `ethics-guardrail.md` §3 (the canonical pair the parity gate checks) |
+| **Bright line / Dark line** | The ethics ship-gate summary; the full guardrail contract is MCP-served and in `ethics-guardrail.md`                           |
+| **Served-by-MCP pointer**   | The `retention_get_mechanic` / `retention_get_ethics_ladder` calls that return the full card                                   |
+| **Motion binding (canon)**  | Motion-bearing cards only: the pinned celebrate-family spring literal `check:motion-contract` keeps undriftable                |
+
+**Full deck field set** — served per card by the Retention Mechanics MCP (not shipped in this
+repo; the maintained deck is versioned per card):
+
+| Field                                             | Meaning                                                                                              |
+| ------------------------------------------------- | ---------------------------------------------------------------------------------------------------- |
+| **Emotional beat**                                | The precise feeling the card engineers, in the user's voice                                          |
+| **Psychology + canonical research**               | Mechanism explanation with named sources; `attribution-uncertain` flags where replication is partial |
+| **Mechanism steps**                               | Numbered implementation sequence                                                                     |
+| **Real app examples**                             | Named apps, named moments, why it works                                                              |
+| **When to use / When NOT to use**                 | Trigger conditions and explicit exclusions                                                           |
+| **Producer recipe**                               | Step-by-step instructions for applying the card to a feature                                         |
+| **Auditor signals — Present / Missing / Misused** | Three-bucket checklist for review                                                                    |
+| **Measurement events**                            | Named PostHog events, required properties, what each proves                                          |
+| **Mobile implementation + reduced-motion**        | Platform-specific code notes; OS reduce-motion fallback is mandatory                                 |
+| **Bright line / Dark line / Guardrail**           | Ethics contract; deterministic ship-gate                                                             |
+| **Pairs with**                                    | Other cards or references this card composes with                                                    |
+| **11-star level**                                 | Star-ladder position from [`eleven-star-experience.md`](eleven-star-experience.md)                   |
+
+---
+
+## Summary Table
+
+| Card                       | Emotional beat                                                  | Primary research                                              | 11-star level | Dark-line to refuse                                     |
+| -------------------------- | --------------------------------------------------------------- | ------------------------------------------------------------- | ------------- | ------------------------------------------------------- |
+| Commitment                 | Ownership — "I said this matters to me"                         | Cialdini 1984; Locke & Latham 1990; Gollwitzer 1999           | 6–7           | Commitment used to manufacture guilt at paywall         |
+| Variable Reward            | Anticipation-then-surprise — "I wonder what I'll get"           | Skinner 1938; Schultz 1997; Berridge 1996                     | 6–7           | Spend prompt on the same screen as the reveal           |
+| Perceived Effort Delay     | Anticipatory trust — "It's working hard for me"                 | Buell & Norton 2011; Norton, Mochon & Ariely 2012             | 6–7           | Artificial sleep timer with no real computation         |
+| Intent Mirroring           | Being seen — "This product understands what I want"             | Norman 2004; Gollwitzer 1999; Cialdini 1984                   | 7             | Mirror on cancel/unsubscribe flow as retention friction |
+| Endowed Progress           | Momentum — "I'm already partway there"                          | Kivetz, Urminsky & Zheng 2006; Hull 1932                      | 6–7           | Phantom credits with no real product operation          |
+| Peak-End                   | Elation-then-completeness                                       | Kahneman & Fredrickson 1993; Norman 2004                      | 6–7           | Manufactured ranking on day one                         |
+| Streak & Loss Aversion     | Protective urgency — "I can't let this die"                     | Kahneman & Tversky 1979; Thaler 1980; Hull 1932               | 5–7           | Paid-only forgiveness with no free grace period         |
+| Reciprocity                | Surprised gratitude — "They gave me something I didn't ask for" | Cialdini 1984; Eyal 2014; Fogg 2019                           | 6–7           | Gift withholds real value behind paywall                |
+| Identity & Self-Expression | Ownership-pride — "This is mine; it reflects who I am"          | Norman 2004; Cialdini 1984; Norton, Mochon & Ariely 2012      | 7             | Identity anchor held hostage on subscription lapse      |
+| Fresh Start                | Clean-slate optimism — "This is a new chapter"                  | Dai, Milkman & Riis 2014; Gollwitzer 1999                     | 6–7           | Temporal-landmark frame leads directly to paywall       |
+| Mastery & Status           | Earned pride — "I've become someone who is good at this"        | Locke & Latham 1990; Eyal 2014; Deci & Ryan 1985              | 6–7           | Level-up reveal coupled with paywall CTA same screen    |
+| Recovery & Trust Repair    | Relieved loyalty — "They handled that quickly and fairly"       | Kahneman & Fredrickson 1993; Norman 2004; Buell & Norton 2011 | 5–7           | Spend prompt inside failure/grief screen                |
+
+---
+
+## Ethics Ladder
+
+**Variable Reward, Streak / Loss Aversion, Scarcity / Urgency, and Social Proof** are
+HIGH-risk mechanisms per the risk table in `knowledge/experience/ethics-guardrail.md`. Before any of
+these cards ships, the artifact doc (`product/ONBOARDING.md`, `PRODUCT.md`, or `ETHICS.md`) must contain
+an experience-card attestation block with all five HIGH-tier fields filled:
+
+```yaml
+experience_card:
+  id: "<card-slug>"
+  mechanism: "<mechanism name>"
+  applied_to: "<screen or feature name>"
+  star_level: <int>
+  posthog_event: "<primary event name>"
+  bright_line: "<one sentence>"
+  dark_line: "<one sentence>"
+  guardrail: "<one sentence>"
+  user_control_escape_hatch: "<where the user can disable or opt out>"
+  ethics_attestation: "<reviewer name and date>"
+```
+
+Run `npm run check:emotional-design -- --root .` before marking any HIGH-risk card
+launch-ready. The validator enforces that no required field is empty and that no
+`spend_prompt_after_reward` pattern appears in the same screen scope.
+
+All other cards require the four MEDIUM-tier fields: `bright_line`, `dark_line`, `guardrail`,
+and (for Perceived Effort Delay) `effort_truthfulness_attestation`.