Skip to content

therealwestninja/advanced-hypnosis-narration-engine

BranchesTags

Repository files navigation

· Advanced Hypnosis Narration Engine ·

A free, AI-driven, fully browser-native hypnosis experience generator.

Generate personalized induction scripts with AI, hear them narrated by the browser's text-to-speech engine, and experience guided sessions with procedural soundscapes, hypnotic visual effects, and real-time customization — all running locally in your browser.

Live here: https://perchance.org/advanced-hypnosis-narration-engine No account required. No downloads.

What's New

  • Motif Engine — every session picks ONE concrete sensory image from a curated 24-motif corpus (a brass key warm from being held, a lighthouse beam passing and returning, a steady ember beneath grey ash, …) and threads it across the whole session as a deliberate through-line:
    • narration — the AI is instructed to weave the motif image 2–4 times across the phases, lightly in early phases and with more weight at the deepest point
    • visuals — when the visualizer is on auto, the motif's preferred visualizer (mandala / candle / ripples / starfield, depending on the motif) is used, and the visualizer's color is tinted toward the motif's palette with a phase-weighted intensity curve
    • audio cue — an optional soft tone (bell / bowl / chime / drop, matched to the motif's sense) plays once at the deepest phase of the session
    • three independent toggles in Smart Director options let you keep the narrative motif while disabling the cue or visual side
    • rotation memory persists across sessions, so 12 consecutive sessions get 12 different motifs
  • Soundscape coherence layer — every procedural soundscape (rain, ocean, fire, forest, wind, bowls, …) now has:
    • a session breath at 1/240 Hz (one ~4-minute cycle) gently modulating master gain by ±10% — soundscapes no longer hit a static steady state, they slowly swell and recede over a session's arc
    • a motif warmth tilt — warm motifs (ember, brass key, banked fire) shift the soundscape's tone filter toward warmer cutoffs; cool / neutral motifs leave it transparent
  • Public Perchance plugin exports — other Perchance generators can now import HNE as a function library:
    • root.hne.freshSeeds(opts) returns evocative sensory anchor phrases from a 60+ phrase corpus tagged across 12 categories
    • root.hne.freshMotifs(opts) returns full motif objects (image, essence, sense, tags, palette, viz, cue) for use in meditation apps, tarot readers, dream journals, sleep aids, poetry generators
    • root.hne.pickFromBank(opts) — generic rotation utility that applies HNE's tag-filter + exclude-set + cold-start fallback + Fisher-Yates shuffle to caller-provided banks. Use when HNE's bundled meditation-register corpus doesn't fit your project's voice (carnival-mystic fortune teller, noir detective, clinical sleep app, …)
    • all three are tag-filtered and rotation-aware with a cold-start fallback
  • Community language packs — translations live in separate Perchance generators that fork the data pack and translate any subset of its keys (missing keys fall through to English). One-line entry in the hneLanguagePacks list registers a translation with HNE's language picker
  • Community content packs — third-party generators can publish persona / method / preset collections that HNE loads lazily on first use
  • Phase-locked entrainment — binaural beats and isochronic tones ramp frequency per session phase across the α/θ/δ brain-wave map, with 6-second crossfades at each boundary
  • 40 Hz gamma flicker visualizer — alpha-grade entrainment surface with conservative implementation (5–15% alpha modulation, narrow centered pulse, blue palette, voice-coupled) and OS-aware reduced-motion fallback
  • Drag-and-drop phase reordering — grab the ≡ handle on any phase row, with a live timeline view above the list showing proportional phase durations
  • Per-phase duration estimate + timeline — every row shows ⏱ Xm Ys inline; click any timeline block to jump to that phase
  • 2D Director calibration pad — combined intensity × pace control on a drag-snap grid
  • A/B experiment mode — blinded sessions on two random presets, rate 1–5, rolling per-preset averages
  • Conversational free-text check-in — type how you feel ("tired but mind is racing") and a lexicon-based parser maps it to mood / tension / energy values
  • Phase audition — 🔊 button on any phase row plays it aloud using current voice settings before committing to a full session
  • Annual review screen — opens after 30+ days of practice; sessions, minutes, completion %, longest streak, mood arc, top guides / methods / contexts. Pure local aggregation, no AI
  • Audio/viz safety relaxation gate — set your profile age range and the overcautious gates step back: gamma flicker confirmation skips, prefers-reduced-motion auto-degrade skips, speech rate caps widen, soundscape volume caps lift. Casual users keep all original safeguards
  • Shareable links — post a persona or a full session config as a real URL with tailored link previews in Discord / Slack / iMessage
  • Streaming script generation — each phase appears live as the AI writes it, instead of a spinner
  • Semantic memory retrieval — the AI recalls past sessions that are relevant to your current intention, not just the most recent ones
  • Regeneration diffs — retry a single phase and pick between the original and the new version side-by-side
  • Token-aware prompt assembly — long profiles, memory, and affirmations never silently exceed the context window
  • Multi-provider AI — use the free built-in Perchance model, or bring your own Anthropic / OpenAI key

What It Does

6-Step Guided Workflow

  1. Choose a Guide 38 built-in personas across four temperaments, each with distinct tone and delivery. Create unlimited custom personas. Import personas others have shared.

  2. Choose a Method 37 induction techniques across therapeutic, somatic, demonstration, rapid, and advanced categories.

  3. Set an Intention Define what you want to achieve. The AI converts this into structured hypnotic suggestions.

  4. Affirmations Add your own affirmations. These are no longer static — the AI:

    • weaves them into sessions
    • structures them across phases
    • adapts delivery style (woven / pulses / light touch)

  5. Configure & Review

    • session length
    • narrator voice
    • soundscape + visuals
    • edit and rearrange generated script
    • adjust phase structure

  6. Generate & Begin Run a fully narrated session with real-time controls.

Quick-Start Presets

31 one-click curated presets — Sleep, Quick Calm, Deep Focus, Pain Relief, Micro Reset, Anxiety Relief, Confidence, Peak Flow, Habit Change, Creative Flow, Energy Boost, Nightcap, Train Journey, Temple Sit, Mission Brief, Goal Lock-in, …

Multi-Day Structured Programs

9 progressive programs that sequence sessions across days/weeks with a coherent arc:

  • 7-Day Sleep
  • 14-Day Calm
  • 10-Day Confidence
  • 6-Week Pain
  • 7-Day Focus
  • 7-Day Depth
  • 5-Day Stress Reset
  • 7-Day Sensuality
  • 5-Day Surrender

Session Model

Sessions are dynamic, not fixed to 5 phases. The full arc includes:

  • settling
  • induction
  • deepener
  • multiple work blocks (up to 5 for deep sessions)
  • affirmation phases (dynamic)
  • lightener
  • reintroduction

Shape: descent → work → recovery → return.

The Motif Engine quietly anchors all of these to a single sensory image, so the session reads as one continuous thing rather than a list of phases.

AI Generation

  • Streaming per-phase generation — each phase renders live with an inline preview you can read as it builds
  • Motif through-line injection — the chosen session motif is part of the shared prompt context for every phase, alongside profile, persona, method, freshness anchors, and memory
  • Fresh metaphor anchors — 4 randomly-picked phrases per session from a 60+ phrase corpus, tag-matched to your profile, injected as "optional anchors" the AI may use to break mode-collapse
  • Semantic memory retrieval — the generator embeds your past session digests and pulls in the most relevant ones when writing a new script (not just the most recent)
  • Token-aware prompt assembly — profile, adaptive context, motif, freshness, and memory blocks are prioritized and trimmed against the context budget
  • Adaptive intelligence — learns what works for you (personas, methods, lengths, depth preferences) and shifts recommendations over time
  • 3-tier hierarchical memory — per-session digests, rolling patterns (every 5 sessions), long-term profile (every 15)
  • Per-phase regeneration with diff — unhappy with one phase? Retry just that one and compare old vs. new side-by-side
  • Multi-provider routing — default free Perchance AI, or plug in your own Anthropic / OpenAI key (the key lives in the browser only)
  • AI scene directives — the AI can emit [BG:rain] / [VIZ:mandala] sentinels mid-script to shift the soundscape or visualizer when the language calls for it (opt-out toggle)

Script Editing (Step 5)

  • drag-and-drop phases via the ≡ grab handle (or ↑/↓ buttons)
  • see a proportional timeline visualization above the list — click any block to jump
  • see estimated duration (⏱ Xm Ys) on every phase row, plus session total
  • audition any phase aloud (🔊 button) before committing
  • remove or add phases (induction, deepener, level, work, affirmation, lightener, wake, reintroduction, sounding, custom)
  • edit full script text
  • regenerate individual phases and compare outputs side-by-side
  • author custom blocks, then ✨ Reimagine them in your guide's voice

This is a core feature, not just a preview step.

Audio

  • 27 procedural soundscapes — fully synthesized via Web Audio (no audio files shipped): rain, ocean, fire, forest, stream, wind, thunder, storm, drone, bowls, heartbeat, binaural, isochronic, chimes, pink/brown/white noise, fan, cave, crickets, temple, spaceship, underwater, clock, train, city, owls, whales
  • Session breath layer — slow 1/240 Hz modulation gives every soundscape gentle long-arc evolution instead of a static steady state
  • Motif warmth tilt — soundscape tone subtly tracks the active motif's color temperature
  • Phase-locked entrainment — binaural beats and isochronic tones ramp their frequency per session phase across the α/θ/δ map, with 6-second crossfades at each phase boundary
  • Narration ducking — soundscape gain dips automatically during speech
  • Voice selection with quality sorting — neural / premium voices ranked first
  • Live adjustment during session — volume, pitch, rate, tempo all editable mid-session
  • Mobile lock-screen media controls — pause / resume from the lock screen

Visuals

  • 30 animated focus effects — spiral, tunnel, vortex, pendulum, mandala, candle, flowfield, plasma, aurora, lissajous, sacred-geo, kaleidoscope, waveform, fireflies, ripples, breathing, ink, smoke, starfield, colorwash, moire, highway, rain, pulse, pocket-watch, spinning-coin, hand-pendulum, canvas-pulse, gamma-flicker, and more
  • Motif palette tint — visualizer color blends toward the motif's secondary palette tone, weighted by phase intensity (faint during settling/induction, peak at deepener/work, fading on lightener)
  • Motif viz biasing — when the visualizer is on auto, the motif's preferred viz list is preferred over the method default
  • 40 Hz gamma flicker — alpha-grade, conservative implementation, voice-coupled
  • Real-time tuning — speed, intensity, complexity all editable mid-session
  • Phase-reactive pacing — visual rate shifts per phase
  • Reduced-motion support — OS preference auto-degrades intense visualizers
  • Background image generation — AI text-to-image, static URL, or YouTube video

Session Player

  • dynamic phase playback (not fixed count)
  • progress tracking + transcript view
  • live settings panel (audio + voice + visualizer + tempo all editable mid-session)
  • emergency exit + grounding system
  • fullscreen + wake lock
  • optional ambient audio recording

UI / UX

  • mobile-first layout (portrait optimized)
  • no horizontal overflow
  • responsive grid-based UI
  • improved readability (contrast + font sizing)
  • clear navigation (back buttons, step clarity)
  • onboarding tour for new users
  • toast notifications (non-blocking, non-interrupting)
  • command palette (Cmd/Ctrl-K)

Data & Systems

History

Stores per session:

  • full script
  • phase structure
  • affirmations + mode
  • profile snapshot
  • session metrics (depth, mood delta, completion state)
  • active motif id (so a session's through-line is searchable in history)

Profiles

User profile is injected into AI generation:

  • goals
  • preferences
  • traits
  • appearance (for personalized imagery)
  • category-scoped sharing controls (what the AI is allowed to see)

Adaptive System

Tracks and updates automatically:

  • completion rates
  • mood-delta variance
  • method/persona preferences by reward signal
  • novelty-seeking vs. repetition patterns
  • depth tolerance
  • recovery mode when sessions trend poorly

Tracking

Expanded analytics:

  • session outcomes
  • usage patterns
  • feature interaction
  • structural preferences

Stats & Charts

  • usage breakdowns (guides, methods, intentions)
  • time-of-day patterns
  • mood change tracking
  • completion rates
  • feature usage maps
  • activity heatmap (GitHub-style)
  • pre/post improvement sparkline

Backup & Sharing

Local

  • full backup / restore (gzipped JSON)
  • script export to .txt
  • compatibility with older formats (automatic migration)

Shareable Links

  • session configs — share persona, method, length, soundscape, affirmations, and optionally the generated script as a single URL
  • custom personas — share a persona you've built; recipients can add it to their library with one click
  • rich link previews — shared URLs produce tailored titles and descriptions when pasted in chat apps
  • community persona packs — opt-in third-party collections load lazily when selected

Privacy-aware share flow

  • recipients see an age-gate prompt before importing personas with potentially sensitive content
  • URL is cleaned after import so a reload doesn't re-trigger the flow

Use HNE as a Perchance Plugin

Other Perchance generators can import HNE's public functions:

hne = {import:advanced-hypnosis-narration-engine}

Then in the consuming generator:

// Pull 3 evocative sensory anchors matching a profile tag set,
// avoiding any recently-used ones.
let anchors = root.hne.freshSeeds({
  profileTags: ["sleep", "release"],
  n: 3,
  excludeSet: ["the slow pull of evening tide"],
});

// Pull ONE recurring session motif — a structured image object with
// palette/viz/cue metadata for threading across narration, visuals, and audio.
let [motif] = root.hne.freshMotifs({
  profileTags: ["focus", "trance"],
  n: 1,
  excludeSet: previouslyUsedIds,
});
// motif = { id, image, essence, sense, tags, palette, viz, cue }

Register caveat: HNE's bundled corpus is meditation/hypnosis-coded — grounding, somatic, kinesthetic ("a held shoulder finally dropping", "a lighthouse beam, passing and returning"). If your project's voice is different — carnival-mystic, clinical, noir, playful — using the curated phrases as primary content will dilute your tone.

For different voices, use pickFromBank instead: it exposes HNE's rotation machinery (tag-filter, exclude-set, cold-start fallback, Fisher-Yates shuffle) operating on your own data:

const fortuneMotifs = [
  { card: "Queen of Cups", tags: "love|intuition",
    image: "tea leaves settling into the shape of a bird" },
  { card: "Ace of Cups",   tags: "love|beginning",
    image: "a chalice brimming, water trembling at the rim" },
  // …
];
const [draw] = root.hne.pickFromBank({
  bank:         fortuneMotifs,
  profileTags:  ["love"],
  n:            1,
  excludeSet:   recentlyDrawn,   // array of card names
  excludeField: "card",          // dedupe key on each item
});
// → { card: "Queen of Cups", tags: "love|intuition", image: "..." }

pickFromBank accepts plain string banks too — useful when you just want HNE's anti-repetition shuffle without object plumbing. See the top-editor docs block for the full signature.

Useful for any project that needs sensory imagery with rotation — meditation apps, tarot readers, dream journals, sleep aids, poetry generators, noir adventure games, anywhere you want variety-with-memory but with your own voice.

To embed the whole HNE UI in another page, use a standard iframe — session links and shared configurations work via URL hash params.

Translation / Language Packs

The English data pack lives in a separate Perchance generator (advanced-hypnosis-narration-engine-data). To translate HNE:

  1. Fork the data generator
  2. Translate any subset of keys (BUILTIN_PERSONAS, METHODS, PRESETS, TOUR_STEPS, GLOSSARY, PROFILE_FIELD_SCHEMA, …); missing keys fall back to the English baseline
  3. Keep the shape identical
  4. Submit a one-line entry to hneLanguagePacks and your translation appears in HNE's language picker for everyone

AI prompt strings (systemPrompt, guidance, methodInstructions) determine how the model narrates — translate them carefully and test the output.

Privacy

  • runs locally in the browser
  • no default data collection
  • no telemetry

Optional (Opt-in Only)

Users may enable:

Remote backup + processing

  • requires explicit consent
  • disabled by default
  • clearly disclosed before activation

Safety

  • contraindication screening
  • session limits + cooldowns
  • grounding system
  • emergency exit
  • alertness check
  • crisis resources
  • reduced-motion defaults for seizure-risk visuals

Accessibility

  • keyboard controls
  • reduced motion support
  • improved contrast + readability
  • larger tap targets
  • focus indicators
  • ARIA labels on navigation

Technical Stack

  • Perchance (split across three generators: main app, data pack, optional community/language packs)
  • ai-text-plugin for streaming generation
  • text-to-image-plugin for backgrounds
  • upload-plugin for shareable content
  • dynamic-import-plugin for community / language packs
  • favicon-plugin, bug-report-plugin, comments-plugin
  • Web Audio API (procedural synthesis throughout — no audio files shipped)
  • Canvas 2D rendering
  • IndexedDB storage (persisted via navigator.storage.persist())
  • Web Speech API (TTS)
  • Wake Lock, Media Session, Vibration, Clipboard APIs

Browser Support

  • Chrome / Edge: full support
  • Safari: full support
  • Firefox: partial (speech-synthesis differences)
  • Mobile: fully supported (lock-screen controls on iOS 16+ / Android 10+)

Features that gracefully degrade when unavailable:

  • semantic memory falls back to recency-based retrieval
  • streaming falls back to full-phase render
  • share links fall back to hash-URL encoding if upload service is unreachable
  • haptics silently skip on desktop
  • motif audio cue silently skips when WebAudio is blocked

Installation

No installation required. Runs entirely in browser.

License

MIT

Credits

Created by therealwestninja Perchance platform: perchance.org

About

Infinitely-customizable five-phase Cognitive Behavioral Therapy (CBT). 25 character voices, 37 clinically-documented induction techniques, text-to-speech narration, procedural soundscapes, hypnotic visual effects, real-time controls, clinical safety protocols, structured multi-day programs, evidence-based outcome tracking. Free & private

perchance.org/advanced-hypnosis-narration-engine

Topics

tracking mindfulness cognitive-science meditation therapy hypnosis daily-tasks cbt mind-control meditation-practice meditation-app wellness-app

Resources

Readme

License

MIT license

Security policy

Security policy
Activity

Stars

1 star

Watchers

0 watching

Forks

0 forks
Report repository

Releases

No releases published

Packages

 
 
 

Contributors