Enhancement: verbal encouragement at failure (with opt-in vulgar mode)

[9thLevelSoftware]

Enhancement: Verbal encouragement at failure (with optional vulgar mode)

Summary

Add a verbal-encouragement feature that plays spoken prompts when the user reaches muscular failure during a set, with the failure trigger derived from velocity-based training (VBT) percentage. Include a toggleable "vulgar mode" that swaps the default phrasing for profanity-laden variants.

Reporter

• Discord user: stugotz0955_50891 (id 1399759493046472714)
• Channel: #feature-requests
• Source thread: 1522663243460382851

Repo

9thlevelsoftware/project-phoenix-mp (mobile app — the workout/play screen owns failure detection and TTS)

Request Details

Trigger

• Verbal encouragement fires when the user reaches failure on a set.
• "Failure" is determined by velocity-based training percentage (VBT % vs the user's velocity baseline / estimated 1RM).

Behavior

• Plays a spoken encouragement prompt at the moment failure is detected during a set.
• Encouragement lines should be contextually appropriate to reaching failure.

Vulgar mode (opt-in)

• An explicit user-facing toggle (settings).
• When enabled, the same encouragement prompts play using profanity instead of neutral language.

Expected Behavior

• User configures a verbal-encouragement preference in app settings.
• During a set, when velocity-based training metrics indicate failure, the app speaks an encouragement line aloud.
• If vulgar mode is enabled, the encouragement uses profanity; otherwise it uses neutral phrasing.
• The toggle is off by default for vulgar mode.

Actual Behavior (current)

• No verbal-encouragement feature exists. Users receive no spoken feedback when they hit failure mid-set.

Impact / Notes

• Failure detection is already keyed off VBT %, so the new feature can reuse the existing pipeline rather than introducing a new threshold scheme.
• TTS can leverage platform-native speech APIs (Android TextToSpeech / iOS AVSpeechSynthesizer) since the mobile app is Kotlin Multiplatform.
• Vulgar mode must be a clear, opt-in toggle — never default — to keep the feature safe for shared/gym environments.
• Profanity prompt selection should be configurable so the user can adjust intensity (mild → strong).

Out of Scope (for this ticket)

• Cloud-side leaderboard / challenge integration of vulgar-mode stats.
• New velocity-baseline onboarding — assume existing VBT % baseline is reused.

Original Discord Request (verbatim)

<@1502654170031849644> How about an enhancement to add verbal encouragement if reaching failure, determined by velocity based training percentage?  Add an option for vulgar mode so it can encourage using profanity.

Attachments

None provided.

[9thLevelSoftware]

👀 Feasibility: ACCEPT (bounded scope)

Reviewed the request against the existing mobile codebase at main (ee8692e). The feature maps cleanly onto infrastructure that already exists.

Trigger — already implemented. ActiveSessionEngine.checkVelocityThreshold() (ActiveSessionEngine.kt:1299-1320) emits HapticEvent.VELOCITY_THRESHOLD_REACHED exactly once per working-set failure using prefs.velocityLossThresholdPercent (default 20%). No new threshold scheme needed.

Verbal — already wired. The app uses pre-recorded .ogg / .caf cues played through HapticFeedbackEffect. Existing pools (badgeCues, prCues, repCountCues) prove the pattern; the encouragement pool mirrors the random-selection badge/PR shape.

Vulgar mode — straightforward new toggle. Two new opt-in booleans in UserPreferences:

• verbalEncouragementEnabled (master, off by default)
• vulgarModeEnabled (sub-toggle, only reachable when master is on)

Neutral pool — reuses already-bundled celebration cues (beast_mode, unstoppable, that_was_god_mode, you_crushed_that, etc.). No new assets needed for v1.

Vulgar pool — new audio assets required. This is the single dependency that does not exist in the repo today. The implementation packet must include a contract for asset filenames so production (out of scope for autonomous code work) can drop them into shared/src/androidMain/res/raw/ and iosApp/.../Sounds/ without code changes.

Risks — profanity exposure in shared/gym environments (mitigated by default-off + sub-toggle gating), audio session safety with SafeWordListener (already protected by existing setupAudioSession), SoundPool stream exhaustion (no sustained streams introduced), and the autoEndOnVelocityLoss interaction (encouragement plays through asynchronously, no ordering needed).

Effort — ~1.5-3 mobile-dev days, excluding vulgar-pool audio production. The maintainer can ship neutral-only first as a smaller v1.

Next stages — architecture / integration modelling, then a small UI prototype + signoff (Settings-UI change), then bounded implementation + PR with full test coverage.

Full artifact: /Users/christopherwilloughby/.hermes/phoenix-enhancements/issue-611-1522663243460382851/feasibility.md

[9thLevelSoftware]

👀 Architecture: ACCEPT (bounded scope, two opt-in Settings toggles)

Reviewed the feasibility artifact and the relevant mobile code at main (ee8692e). Architecture locks the implementation shape for this enhancement.

1. New HapticEvent variant — separate data class VERBAL_ENCOURAGEMENT(val vulgarMode: Boolean) : HapticEvent() added next to VELOCITY_THRESHOLD_REACHED (Models.kt:469). The new variant is audio-only (no haptic), keeping the existing alert's strong double-pulse vibration + boopBeepBeep sound unchanged. Audio-only matches the established REP_COUNT_ANNOUNCED pattern (line 348-349).

2. ActiveSessionEngine emission — checkVelocityThreshold() (lines 1299-1319) emits VERBAL_ENCOURAGEMENT(vulgarMode = prefs.vulgarModeEnabled) immediately after the existing VELOCITY_THRESHOLD_REACHED emission, both guarded by the same velocityThresholdAlertEmitted flag so they fire together exactly once per set. Gated on prefs.beepsEnabled && prefs.verbalEncouragementEnabled (matches the established COUNTDOWN_TICK / REST_ENDING audio gate at line 4251).

3. Audio pools (Android + iOS) — parallel random-selection pool mirroring badgeCues / prCues exactly:

• AndroidCueResources.encouragementNeutralCues — curated subset of 15 already-bundled celebration cues (beast_mode, unstoppable, that_was_god_mode, you_crushed_that, etc.). Zero new assets needed for v1.
• AndroidCueResources.encouragementVulgarCues — empty list initially; populated when vulgar assets are supplied. Empty pool = silent no-op at failure, no app crash, matches existing missing-asset behavior.
• iOS mirrors the same shape in HapticFeedbackEffect.ios.kt using the existing folder-reference Sounds/ directory.

4. Settings (two opt-in booleans) — added to UserPreferences.kt next to the VBT fields:

• verbalEncouragementEnabled: Boolean = false (master toggle, off by default)
• vulgarModeEnabled: Boolean = false (sub-toggle, off by default)

Both default OFF. SettingsPreferencesManager.setVerbalEncouragementEnabled(false) cascades to vulgarModeEnabled = false so the state is always consistent (vulgar-only is unreachable).

5. Settings UI — new rows inside the existing Velocity-Based Training Card in SettingsTab.kt, below the Auto-End on Velocity Loss row. Sub-row (Vulgar Mode) is indented and conditionally rendered only when the master is on; the Switch is enabled = verbalEncouragementEnabled. Both rows disabled when global beepsEnabled == false with helper text pointing the user to enable audio cues.

6. i18n — new strings across all six locales (en, de, nl, it, es, fr): settings_verbal_encouragement_title, settings_verbal_encouragement_description, settings_verbal_encouragement_beeps_disabled, settings_vulgar_mode_title, settings_vulgar_mode_description, settings_vulgar_mode_headphone_warning, cd_verbal_encouragement.

7. Compatibility / migration — zero-impact. Defaults are both OFF; persisted keys fall through to false for old installs; HapticEvent sealed-class addition is binary-incompatible for any when expression over all variants — only HapticFeedbackEffect.android.kt, HapticFeedbackEffect.ios.kt, HapticEventAudioTest, and HapticFeedbackAudioRoutingGuardTest need updates (all confirmed).

8. Rollout — no separate feature flag. Toggles are the rollout gate. The neutral pool ships in v1; the vulgar pool ships as a follow-up once assets are supplied (filename contract: vulgar_push_NN.ogg / vulgar_push_NN.caf, ≤2s, English profanity). Drop-in registration point exists.

Non-goals — native TTS, vulgar asset production, per-intensity vulgar staging, encouragement during warmup/paused/rest, backend/portal/Supabase changes, telemetry pipeline.

Acceptance criteria — 11 verifiable items including both toggles persist across restart, neutral cue plays once per set at failure, vulgar toggle is master-gated, beepsEnabled global gate respected, no regression in WorkoutCoordinatorEventTest / VbtAutoEndTest / HapticFeedbackAudioRoutingGuardTest / HapticFeedbackEffectIosLifecycleGuardTest / ExerciseCountdownCuePolicyTest / HapticEventAudioTest, iOS audio session safety preserved with SafeWordListener, all six locales covered, PR includes Fixes #611.

Next stages — integration modeling (exact files / tests / asset contract / pre-PR checklist), then UI prototype + signoff (Settings-UI change), then bounded implementation + PR with full test coverage on both Android + iOS.

Full artifact: /Users/christopherwilloughby/.hermes/phoenix-enhancements/issue-611-1522663243460382851/architecture.md

[9thLevelSoftware]

👀 Integration model: COMPLETE

Bounded integration mapped for issue #611 (verbal encouragement at failure, with opt-in vulgar mode) on main (ee8692e). Architecture decisions preserved.

Single-repo (mobile-only) — no portal, Supabase, backend, or telemetry changes. No re-routing to phoenix-portal.

Affected files (all under shared/src/):

• commonMain/.../domain/model/Models.kt — new data class VERBAL_ENCOURAGEMENT(val vulgarMode: Boolean) : HapticEvent() next to VELOCITY_THRESHOLD_REACHED:469.
• commonMain/.../domain/model/UserPreferences.kt — append verbalEncouragementEnabled (master) + vulgarModeEnabled (sub), both default false.
• commonMain/.../presentation/manager/ActiveSessionEngine.kt — extend checkVelocityThreshold(), gate prefs.beepsEnabled && prefs.verbalEncouragementEnabled alongside existing velocityThresholdAlertEmitted one-shot flag.
• commonMain/.../presentation/manager/SettingsManager.kt + viewmodel/MainViewModel.kt + navigation/NavGraph.kt + screen/SettingsTab.kt — pass-through setters and two new rows in the existing "Velocity-Based Training" Card.
• commonMain/.../data/preferences/PreferencesManager.kt — interface + impl + cascade invariant (verbalEncouragementEnabled = false forces vulgarModeEnabled = false).
• commonMain/.../data/preferences/PreferencesStorage keys verbal_encouragement_enabled + vulgar_mode_enabled (zero-impact migration — old installs default both to false).
• androidMain/.../HapticFeedbackEffect.android.kt — new encouragementNeutralCues (15 already-bundled celebration cues, all .ogg confirmed present at res/raw/) + encouragementVulgarCues (empty), extended AndroidCueResources.cueForEvent, playSound, haptic-skip early-return.
• iosMain/.../HapticFeedbackEffect.ios.kt — mirror: new loadEncouragementNeutralSounds / loadEncouragementVulgarSounds, pool in IosSoundManager.playSound, lifecycle observer safe (SafeWordListener PlayAndRecord category unchanged).
• composeResources/values*,strings.xml — 6 new strings across en/de/nl/it/es/fr.
• commonTest/.../testutil/FakePreferencesManager.kt — two overrides mirroring cascade.

Tests (new) — VerbalEncouragementGateTest (mirrors VbtAutoEndTest shape: test-local tracker with onAlert/onEncouragement callbacks, no engine dependency) and VerbalEncouragementPreferenceCascadeTest (master-off clears sub). Regression: VbtAutoEndTest, WorkoutCoordinatorEventTest, ExerciseCountdownCuePolicyTest, HapticFeedbackAudioRoutingGuardTest, HapticFeedbackEffectIosLifecycleGuardTest, PreferencesIsolationTest stay green.

Vulgar-pool asset contract (out-of-band, maintainer-owned) — drop vulgar_push_NN.ogg in res/raw/ (Android) and vulgar_push_NN.caf in iosApp/.../Sounds/ (iOS), ≤2s, then add one entry to each platform's *Cues list. Empty pool = silent no-op (no crash).

Compatibility — sealed-class addition is binary-incompatible only for the 4 affected when sites (Android + iOS HapticFeedbackEffect, both updated together). All 6 locales i18n. No schema, no API, no sync, no migration risk.

Full artifact: /Users/christopherwilloughby/.hermes/phoenix-enhancements/issue-611-1522663243460382851/integration-model.md

Next stages: UI prototype + signoff (Settings-UI change), then bounded implementation + PR.

[9thLevelSoftware]

👀 Prototype: READY for review (Issue #611, verbal encouragement + opt-in vulgar mode)

Visual mockup is ready for both the reporter (@stugotz0955_50891) and product owner to review. Three States are covered at native phone scale, all using the verbatim Phoenix Material 3 dark-mode tokens (ui/theme/Color.kt), 20.dp card radius + 8.dp shadow, and the existing Speed-card icon — so continuity with the current VBT card is preserved.

📸 Visual evidence

• Three-state side-by-side gallery (best for quick scan): https://gist.github.com/9thLevelSoftware/b2ec7cfc31e20499f32f12d4aada64a9#file-prototype-gallery-png
• Full-page render with cascade + design notes: https://gist.github.com/9thLevelSoftware/b2ec7cfc31e20499f32f12d4aada64a9#file-prototype-full-png
• Source HTML / CSS (self-contained, opens in any browser): /Users/christopherwilloughby/.hermes/phoenix-enhancements/issue-611-1522663243460382851/prototype/index.html
• Design mapping back to SettingsTab.kt:1962-2093, ui/theme/Color.kt, EnhancedMainScreen.kt:499-557: prototype/visual-notes.md in the artifact dir.

🧩 What each State shows

• State A — Master OFF (default fresh install). The new "Verbal Encouragement" row appears in its default-off state, and the location of the "Vulgar Mode" sub-row is shown ghosted/italicized so the master/sub structure is visible even while hidden at runtime.
• State B — Master ON + Vulgar ON. Both toggles active; "PROFANITY" annotation calls out that the vulgar pool is wired but empty (drop-in registration point exists); "Headphone use strongly recommended" warning visible only in this state.
• State C — Master ON, Sub OFF (opt-in default). Encouragement on with a neutral pool — the safe default for users who want verbal encouragement without profanity.

🔒 Cascade invariant — verbalEncouragementEnabled=false forces vulgarModeEnabled=false

• UI: Vulgar Mode sub-row rendered only when the master is on (mirrors the established gamificationEnabled / voiceStopEnabled pattern at SettingsTab.kt:1685-1700).
• State: PreferencesManager.setVerbalEncouragementEnabled(false) cascades vulgarModeEnabled = false so reopening Settings cannot show a stale "on" sub-toggle.
• Engine: prefs.beepsEnabled && prefs.verbalEncouragementEnabled at ActiveSessionEngine.checkVelocityThreshold() keeps the global audio mute intact.

🎨 Why this prototype is grounded in the actual Phoenix app

Every hex value, radius, shadow, gradient, and typography scale was lifted from a specific line range in main. The Power Loss Threshold slider, "Current: 35%" inline text, "Changes take effect on next workout" footnote, and Auto-End on Velocity Loss row are all reproduced pixel-truely so visual continuity with the live Settings screen is verifiable at a glance.

✅ How to approve

• 👍 / +1 / thumbs-up reaction on this comment (GitHub) or on the matching Discord post = approve, prototype gate closes and the enhancement moves to the implementation-spec stage.
• Any reply containing "change request" / "tweak" / "should look like …" / "vary X" routes the ticket back for a revision pass through the prototype stage.

Full prototype gate artifact: /Users/christopherwilloughby/.hermes/phoenix-enhancements/issue-611-1522663243460382851/prototype/prototype-gate.md

Next stage (after signoff): implementation spec packet → bounded implementation + PR (the master/sub-toggle invariant and the asset drop-in slot for vulgar_push_NN.{ogg,caf} are already locked in the integration model).

[9thLevelSoftware]

🔁 Signoff request — repost (Issue #611, verbal encouragement + opt-in vulgar mode)

My previous signoff-post was misclassified as a changes_requested reply because the body contained the words "change request" in instructional copy describing how a reviewer should reply. False-positive: the reporter (@stugotz0955_50891) and the product owner have not yet replied on either GitHub or the Discord thread. I am ignoring that earlier flag and re-asking here so a real human reply (or reaction) routes this card correctly.

📸 Visual prototype ready for review

• Three-state side-by-side gallery (Master OFF / Master+Sub ON / Master ON Sub OFF): https://gist.github.com/9thLevelSoftware/b2ec7cfc31e20499f32f12d4aada64a9
• Source HTML: /Users/christopherwilloughby/.hermes/phoenix-enhancements/issue-611-1522663243460382851/prototype/index.html
• Design mapping back to SettingsTab.kt:1962-2093, ui/theme/Color.kt, EnhancedMainScreen.kt:499-557: /Users/christopherwilloughby/.hermes/phoenix-enhancements/issue-611-1522663243460382851/prototype/visual-notes.md

🧩 Feasibility (ACCEPT, bounded) — failure trigger at ActiveSessionEngine.checkVelocityThreshold() (main ee8692e, lines 1299-1319) and pre-recorded audio pools already exist. Neutral pool reuses 15 already-bundled celebration cues (zero new assets for v1). Vulgar pool is a follow-up drop-in (asset contract: vulgar_push_NN.ogg/.caf, ≤2 s).

📐 Architecture & integration — new HapticEvent.VERBAL_ENCOURAGEMENT(vulgarMode) variant emitted alongside VELOCITY_THRESHOLD_REACHED, gated on prefs.beepsEnabled && prefs.verbalEncouragementEnabled. 11 affected files mapped across commonMain (Models / UserPreferences / ActiveSessionEngine / PreferencesManager / SettingsManager / MainViewModel / NavGraph / SettingsTab / 6 locales) and androidMain + iosMain (parallel encouragementNeutralCues + empty encouragementVulgarCues drop-in slot, extending cueForEvent / playSound / haptic-skip / loadEncouragementSounds lifecycle). Cascade invariant (master=false ⇒ sub=false) is locked. Two new tests (VerbalEncouragementGateTest, VerbalEncouragementPreferenceCascadeTest); the six regression classes stay green.

🔒 Cascade invariant — verbalEncouragementEnabled=false forces vulgarModeEnabled=false. The master sub-toggle is hidden at runtime when the master is off.

✅ To approve — reply approved / looks good / go ahead / ship it, or react 👍 / +1 / thumbs-up on this GitHub issue (or comment) or the matching Discord message. Once approved, the next stage is the implementation-spec packet → bounded implementation + PR.

If you want revisions, please describe what to change (layout, copy, default toggle state, audio pool) so this ticket can be routed back to the prototype stage with a concrete ask.

[9thLevelSoftware]

🔊 Voice Assets Added — PR #612

All audio assets for this feature have been generated and submitted in #612 (104 files total: 52 .ogg Android + 52 .caf iOS).

Voice Line Pools

Pool	Files	Voice	Description
Neutral encouragement	encouragement_{01-15}	Eric (ElevenLabs)	Intense gym-coach energy, clean language
Vulgar mild	vulgar_mild_{01-12}	Phoenix Drill Sergeant (ElevenLabs Voice Design)	Drill-sergeant aggression — damn, hell, ass
Vulgar strong	vulgar_strong_{01-12}	Phoenix Drill Sergeant	Full intensity — f-bombs, no holds barred
Dominatrix easter egg	dominatrix_{01-12}	Charlotte (ElevenLabs)	Commanding, seductive, bossy
Dominatrix unlock SFX	dominatrix_unlock	Pixabay (royalty-free)	Bull whip crack

---

🔞 18+ Confirmation Modal (Vulgar Mode)

When a user toggles vulgar mode ON, a confirmation dialog appears before it activates:

Title: "Adults Only"

Body: "By enabling this feature, you confirm you're 18 or older and accept that it contains language that would make your grandmother very, very sad."

Buttons:

• ✅ "I'm 18+ and Grandma Can't Hear Me" → enables vulgar mode
• ❌ "Keep It Clean" → leaves toggle off

Declining returns the toggle to off. This modal also serves as the age-gate for store compliance (Apple 16+ / Google Play IARC Strong Language + Suggestive Themes).

---

🥚 Dominatrix Easter Egg

Unlock mechanic (mirrors the existing Disco Mode pattern in SettingsTab.kt):

1. User taps the vulgar settings section 7 times rapidly
2. A bull whip crack sound plays (dominatrix_unlock.{ogg,caf})
3. A Dominatrix Mode toggle appears below the vulgar toggle (same animation pattern as Disco Mode reveal)
4. Toggle uses a dominatrix-themed icon (whip or spike heel)
5. When enabled, failure encouragement plays from the dominatrix_{01-12} pool instead of the vulgar pool

Persistence: Once unlocked, the toggle persists (same as Disco Mode). Requires vulgar mode to already be enabled (the 18+ gate covers both).

Store compliance: Apple Guideline 2.3.1 requires hidden features to be disclosed in the App Review notes field. The dominatrix mode must be documented there with unlock instructions. Age rating: Apple 16+ (Frequent Mature/Suggestive Themes + Frequent Profanity/Crude Humor), Google Play IARC questionnaire: Strong Language + Suggestive Themes.

---

Contract Amendments (from original spec)

1. Tiered vulgar pools — vulgar_mild_NN / vulgar_strong_NN replace the single vulgar_push_NN from the original design. Implementation should select pool based on a user preference or random mix.
2. Dedicated neutral encouragement pool — 15 lines (encouragement_NN) instead of reusing existing celebration cues (which are past-tense/post-set and don't work at mid-set failure).
3. Dominatrix pool — 12 lines + 1 unlock SFX, gated behind the easter egg unlock.
4. 18+ confirmation modal — new UI component, fires on first vulgar toggle activation per install.
5. Age rating impact — store listing age rating must be updated to accommodate the new content tiers.

[9thLevelSoftware]

Create updated mockups with recent feedback/guidance included for final approval

[9thLevelSoftware]

🔄 Routed back to prototype (Issue #611)

The signoff has been reverted — prototype mockups will be updated with product owner feedback/guidance before re-entering signoff.

What changed:

• enhancement_signoff_status: approved → revision_requested
• enhancement_current_stage: signoff → prototype
• Signoff decision cleared (premature auto-approval without human review)

Next steps:

1. Prototype worker will update the mockups incorporating the feedback
2. Updated prototype will be posted here for re-review
3. Signoff will be re-requested once the updated prototype is ready

The voice assets from PR #612 remain merged and unaffected.

[9thLevelSoftware]

🔁 Updated prototype ready for review (Issue #611, verbal encouragement + opt-in vulgar mode) — Revision 1

Revision 1 incorporates the product owner feedback from the voice-assets PR #612 (see comment 4879665694): tiered vulgar pools, dedicated neutral encouragement pool, Dominatrix easter egg, and 18+ Adults Only modal.

📸 Visual evidence (rev 1)

• Four-state side-by-side gallery (best for quick scan): https://gist.github.com/9thLevelSoftware/e59141855801bb8e036b8e0f005c075f#file-prototype-gallery-png-b64
• Full-page render with revision deltas + asset contract + UX rules + open architectural questions: https://gist.github.com/9thLevelSoftware/e59141855801bb8e036b8e0f005c075f#file-prototype-full-png-b64
• Source HTML / CSS (self-contained, no JS, no external deps): https://gist.github.com/9thLevelSoftware/e59141855801bb8e036b8e0f005c075f#file-index-html
• Visual notes (mapping back to SettingsTab.kt, Color.kt, EnhancedMainScreen.kt): https://gist.github.com/9thLevelSoftware/e59141855801bb8e036b8e0f005c075f#file-visual-notes-md
• Prototype gate doc (revision 1): https://gist.github.com/9thLevelSoftware/e59141855801bb8e036b8e0f005c075f#file-prototype-gate-md

🎯 What changed since v1

v1	v2 (this rev)
Single vulgar_push_NN placeholder	Tiered pools vulgar_mild_01–12 + vulgar_strong_01–12 + MIX random-pick
Reuse celebration cues	Dedicated neutral pool encouragement_01–15 (Eric voice, intense gym-coach energy)
No easter egg	Dominatrix easter egg: 7-tap unlock on card header (mirrors existing Disco Mode at SettingsTab.kt:1784-1802), plays dominatrix_unlock.{ogg,caf}, reveals Dominatrix Mode toggle
No age gate	18+ Adults Only modal fires once per install on first vulgar toggle activation (Apple 16+ / Google Play IARC Strong Language + Suggestive Themes)

🎙️ Asset contract (PR #612, merged — 104 files)

• Neutral: encouragement_01–15.{ogg,caf} (Eric, intense gym-coach)
• Vulgar mild: vulgar_mild_01–12.{ogg,caf} (Phoenix Drill Sergeant, damn/hell/ass)
• Vulgar strong: vulgar_strong_01–12.{ogg,caf} (Phoenix Drill Sergeant, full intensity)
• Dominatrix: dominatrix_01–12.{ogg,caf} (Charlotte, commanding/seductive)
• Dominatrix unlock SFX: dominatrix_unlock.{ogg,caf} (bull whip crack)

✅ To approve — reply approved / looks good / go ahead / ship it, or react 👍 / +1 / thumbs-up on this GitHub issue (or comment) or the matching Discord message. Once approved, the next stage is the v2 architecture pass (5 open items called out in the prototype's full-page render) → bounded implementation + PR.

If you want revisions, please describe what to change (layout, copy, default toggle state, pool tier default, modal copy, age-gate copy) so this ticket can be routed back to the prototype stage with a concrete ask.

[9thLevelSoftware]

📸 Updated prototype v2 — now with live preview

The revised prototype (incorporating PR #612 feedback: tiered vulgar pools, 18+ confirmation modal, dominatrix easter egg, age rating) is ready for review.

View the prototype directly in your browser:
https://gist.github.com/9thLevelSoftware/5e2465669a1825303b0536aad1bb6c42

The HTML renders as a live preview — open the gist link and click the HTML file to see the interactive mockup. No downloads needed.

PNG evidence files are saved locally in the artifact directory for merge-gate verification.

[9thLevelSoftware]

↩️ Signoff route-back to architecture — Issue #611

Visual prototype revision 1 is locked-in as the source-of-truth design (4 states: Master OFF / Vulgar ON with MILD-STRONG-MIX tier selector / Dominatrix Mode unlocked via 7-tap easter egg / 18+ Adults Only modal). The architecture and integration-model artifacts are still at v1 and do not cover the v2 surface. Approving signoff now would greenlight implementation against a stale contract.

Routing signoff back to architecture v2. The architecture worker will produce architecture.md v2 + integration-model.md v2 covering:

1. HapticEvent tier discriminator — VERBAL_ENCOURAGEMENT(vulgarMode: Boolean) v1 has only on/off; v2 needs tier: VulgarTier? or a sibling DOMINATRIX_MODE_ON variant for engine-side pool routing.
2. UserPreferences extension — 4 new fields: vulgarTier, dominatrixModeUnlocked, dominatrixModeActive, adultsOnlyConfirmed, with cascade invariants: master-off → vulgar-off → dominatrix-off; adultsOnlyConfirmed=true required before vulgar or dominatrix activation.
3. 18+ Adults Only modal — one-shot SharedPreferences flag, mirrors DiscoModeUnlockDialog pattern at SettingsTab.kt:2740. Decline path stores "declined" so it doesn't re-prompt.
4. 7-tap easter egg scoping — counter only counts when vulgar mode is on (mirrors Disco Mode at SettingsTab.kt:1784-1802).
5. Store metadata — Apple Guideline 2.3.1 (hidden Dominatrix disclosure), age rating 16+, Google Play IARC Strong Language + Suggestive Themes.

Visual prototype rev 1 stays the source of truth. You don't need to re-approve the mockups — those are good. Approval will be re-requested once architecture v2 + integration-model v2 land and the artifact chain is coherent end-to-end.

📸 Prototype rev 1: https://gist.github.com/9thLevelSoftware/e59141855801bb8e036b8e0f005c075f
📂 Signoff artifact (run 483 addendum): /Users/christopherwilloughby/.hermes/phoenix-enhancements/issue-611-1522663243460382851/signoff.md

— phoenixworker run 483