# AudioBoost editorial approval workflow

Use this reference when Margarita reviews a long AudioBoost, sends numbered decisions, or asks for an independent Russian-language read-through.

## 1. Review before editing

### One-criterion-at-a-time approval mode

When Margarita asks to review a long AudioBoost against a numbered audit or checklist, present **one criterion per message** and pause for her decision. Do not send the remaining criteria early, even if the full audit is already available.

For each criterion:

1. state the criterion and status;
2. quote only the relevant passages;
3. explain the issue briefly;
4. give a concrete proposed wording when a change is useful;
5. wait for approval;
6. when approved, patch the canonical working script immediately;
7. confirm the exact change, then proceed to the next criterion.

Treat short replies such as `ок`, `да`, `давай меняем`, `как ты сказала` as approval of the immediately preceding recommendation. Preserve explicit `оставить` decisions. This sequential mode is preferred because it lets Margarita examine the language closely and co-author each phrase without being overloaded by the whole audit.

When asked to “перепроверь и пришли странные места”, do **not** silently rewrite the script. Return a prioritized list containing:

1. exact quotation;
2. why it is awkward in spoken Russian (grammar, calque, unclear referent, mixed metaphor, overload, difficult oral rhythm);
3. one or two concrete alternatives.

Separate clear errors from optional stylistic refinements. Margarita decides which unusual images remain.

### Serial review mode for multi-criterion audits

When Margarita asks to review an AudioBoost by a numbered checklist, present **one criterion at a time** rather than sending the whole audit at once. For each criterion:

1. show its number and exact question;
2. give a concise status and only the relevant quotations;
3. explain the issue and offer concrete wording if a change is warranted;
4. stop and wait for Margarita’s decision;
5. if she approves, patch the canonical working script immediately and verify the replacement;
6. mark that criterion approved, then present the next criterion only.

Do not bundle later criteria into the same message or repeat the whole audit after every decision. If voice transcription contains obvious malformed phrases, restore the intended grammar while preserving Margarita’s meaning, then show the exact final wording inserted.

## 2. Applying numbered decisions

- Map every number back to the exact quoted issue; do not rely on number alone when the conversation contains more than one numbered list.
- Preserve explicit “оставляем” decisions.
- If Margarita says “новый вариант”, write a fresh phrase serving the same function rather than recycling a rejected option.
- Apply approved changes to the full script **and** any extracted Hacking Process block.
- After editing, show the changed passages, not only a general summary.

## 3. Canonical Hacking Statement

In every cycle preserve this formula **exactly**, including punctuation, spelling, `setup`, `touchstone`, and final `Lumos!`:

> True and false, soft and hard, upgrade and restart, bootstrap and setup. This is a touchstone where the future is shown. Lumos!

Never append `All corrupted nodes deleted`, `Fragmented fields reintegrated`, or other lines without Margarita’s explicit approval.

## 4. Hacking Process variety and structure

- Each clearing cycle needs a different entrance and a distinct function; do not repeat “во всех временных интервалах / версиях / копиях / измерениях…” in every cycle.
- Source clearing texts supply mechanics only. Rebuild diction, syntax, and rhythm in Cosmos Luminosity language so Access Consciousness is not recognizable.
- Architecture per cycle: detailed old-code recognition/clearing → exact Hacking Statement → **no more than three questions** → Lumos activation.
- Grade cycles and questions from simple/body-accessible toward systemic, future-facing, creator-level expansion.
- Do not erase the useful function of control. Clear anxiety, surveillance, repeated checking, and premature intervention while preserving control as living navigational clarity.

## 5. Discipline thesis that must survive revisions

After every major rewrite, verify that discipline still contains:

- partnership among inner parts;
- an agreement among desire, body, attention, rhythm, present self, and future self;
- contact with what the discipline is for and what future it creates;
- rhythm, cycles, timing, pauses, return, and exact action;
- warmth, play, celebration, coziness, and material embodiment.

Explain `живой материал` on first important use: body, energy, attention, sensitivity, time, experience, desires, and the project/relationship/life area being created.

## 6. Sensory-to-transformation bridge

Do not jump abruptly from cinnamon craft to discipline. Use a gradual bridge:

1. stay with craft precision (hands, timing, drying, distillation, dosage);
2. let one part of the listener keep seeing the sensory scene while another recognizes a personal process;
3. invite the sequence into bodily perception before naming it;
4. let the old meaning of “discipline” recede and a warm living meaning emerge;
5. only then state the new discipline model.

Permissive hypnotic language may use `можно заметить`, `возможно, одна часть…`, `можно пока не называть это словом`, `чем глубже… тем свободнее…`. Keep it transparent, gentle, and non-coercive; prioritize embodied guidance over claims of hidden control.

## 7. Questions resembling Access syntax

Stems such as `Какой энергией я могу быть…`, `Какой ясностью присутствия я могу быть…`, and `Какой добротой я могу быть…` may feel effective to Margarita even while remaining recognizably Access-like. Do not bulk replace them during final script editing when she asks to keep them. Mark them for a separate phrase-by-phrase authorial language session **after the AudioBoost is approved**. Draft alternatives are not canon until approved.

## 8. Spoken-Russian checks

Prioritize:

- clear subjects and referents;
- concrete verbs instead of passive abstractions;
- one coherent metaphor per sentence;
- physically accurate craft descriptions;
- flowing oral rhythm rather than chains of definitions;
- Russian wording where an English tech term breaks the sound texture;
- consistent second-person singular (`твой`, not accidental `ваш`).

Read transitions aloud mentally: a beautiful written phrase may still be difficult to record.

## 9. Marketing passport, authorial opening, and signature ending

A recording-ready AudioBoost needs reusable product positioning before the spoken script and a spoken frame around the journey.

### Marketing passport before the recording script

Keep this material outside `## Полная версия текста для записи`, so it can be used on a landing page, product card, post, or announcement without increasing the recorded duration:

1. a Cosmos Luminosity-style title naming the symbolic material and transformation;
2. one charged subtitle/tagline;
3. `О чём этот AudioBoost` — 2–3 compact paragraphs;
4. `Цель AudioBoost` — one clear transformation statement;
5. `Кому подходит` — concrete audience and use cases;
6. `Кому может не подойти` — honest fit boundaries: duration, trance/imagery preference, listening situation, and no medical or guaranteed-result expectations.

The passport should feel desirable, specific, and premium rather than clinical. When the title changes, synchronize the spoken introduction, preview `<title>`, and meta description. Keep the passport clearly outside the spoken script and do not count it in runtime.

### Opening sequence

1. Margarita briefly introduces herself and Cosmos Luminosity.
2. She names the AudioBoost and the journey it opens using the same approved title as the marketing passport.
3. Explain the format through the **listener’s experience and the route of transformation**, not as a clinical or technical product definition. Avoid stacking abstractions such as `внутренняя архитектура`, `определённая внутренняя функция`, `внутренняя сборка`, and `пространство обновления` in the opening. Bring in the symbolic material and shared journey early.
4. State the specific purpose without revealing or explaining the entire transformation in advance.
5. If terminology is still being authored, distinguish a descriptive phrase from a canonical term. Do not canonize a label simply because it reads well in one introduction.
6. Add a short listening-safety line for driving or attention-demanding activity.
7. In a safe setting, recommend closing the eyes **for the fullness of inner imagery**. Phrase it as Margarita’s recommendation rather than an absolute command; posture may change and the listener need not follow every word.
8. Transition gradually into relaxation: ownership of time → support under the body → shoulders/jaw/hands → one or two breaths → sensory descent.

### Spoken-language approval loop for an opening

When Margarita dislikes an opening, do not patch immediately. Work in short semantic units:

1. identify what sounds official, abstract, passive, generic, or unlike her voice;
2. offer a small set of alternatives for each disputed function (term, mechanism, destination of energy, future-to-matter formulation);
3. when she chooses numbered options, record the selection and preserve other approved options as a reusable variation bank;
4. when the title itself is unsettled, offer a focused naming set organized by dominant promise (symbolic, action-led, perfumery-led, method-led), identify a short top set, and wait for selection;
5. assemble the whole opening in flowing spoken Russian;
6. read for breath, rhythm, repeated abstractions, repeated sentence stems/verbs, and whether Margarita’s companion presence is felt;
7. patch only after she approves the assembled passage. Once approved, synchronize the chosen title across the marketing passport, spoken opening, preview title, and meta description.

After piecemeal approvals or direct insertions, always show the **entire current spoken opening** from author introduction through the first relaxation breath. Check it as one oral unit for duplicate route announcements (`в этом путешествии…`), repeated product explanation, long pre-predicate lists, landing-page language leaking into the recording, and delay before the first body cue. Marketing facts such as perfumery education may appear briefly in the spoken frame, but the richer audience/benefit explanation belongs in the marketing passport outside the recording. Do not call the opening final until this assembled read-through is approved.

Prefer active transformations (`проводят смысл…`, `возвращают энергию…`, `открывают доступ…`) over passive language such as `складываются в пространство`. Do not overclaim automatic effects: preserve agency with `помогают`, `могут`, or a transparent statement of the practice’s intended function.

Do not jump from product explanation directly to deep trance. Use a bridge such as `если пространство уже позволяет расслабиться…`. Reduce performance anxiety with transparent permission: no need to find the “correct” sensation or follow every word.

### Ending sequence

1. Leave one integration question or image.
2. Return attention to support, breath, sounds, fingers/toes, stretching, and sight.
3. Name what the listener can take into the day and one possible exact action.
4. Briefly explain when the AudioBoost may be replayed.
5. Close in Margarita’s voice with **«Вселенски обнимаю»**, followed by a theme-specific blessing rather than a generic sign-off. This phrase is mandatory in every AudioBoost ending.
6. Optionally identify Margarita and Cosmos Luminosity once more.
7. End the complete AudioBoost with the exact signature **`Stay Lumos!`**. This is mandatory branded language; preserve the English spelling and exclamation mark.

The return should be complete enough for a deep audio journey; do not end immediately after the strongest question.

## 10. Hypnotic-language audit

Run the 16-point trance audit in the order defined in `references/audioboost-trance-language.md`. Criterion 16 verifies the author's continuous, varied, implicit companion presence throughout the journey. If Margarita chooses sequential approval, keep a visible criterion number, apply each approved edit before advancing, and do a final whole-file verification only after criterion 16. Do not re-open an approved criterion unless a later change conflicts with it.

When Margarita asks to make the whole script more hypnotic, first distinguish **hypnotic flow** from saturation with hypnosis markers. The goal is a continuous permissive atmosphere with selected constructions at transitions, not a hypnotic phrase in every paragraph.

If she says she will provide a ChatGPT conversation, transcript, course note, or link containing her preferred constructions:

- wait for and inspect that source before editing;
- extract the exact construction families she approved;
- compare them against the script;
- first return `точная цитата → why it feels sharp/technical → proposed softening → construction type`;
- do not silently insert changes until she approves the audit, unless she explicitly asks for direct editing.

Audit especially:

- abrupt jumps between explanation, sensory scene, Hacking Process, and return;
- hard commands where permission would preserve agency;
- dense technical/clearing paragraphs with no breath or sensory bridge;
- repeated use of one pattern (`и пока…`, `можно…`) that makes the trance mechanical;
- embedded suggestions that imply inevitability, hidden control, or guaranteed change.

Useful ethical families may include pacing-and-leading, permissive suggestion, sensory grounding, one-part/another-part attention, temporal linking, gentle presupposition, open-ended imagery, and future pacing. Vary them and keep choice explicit. Hypnotic language must support agency, bodily safety, and transparent purpose.

## 11. Duration estimation

Estimate only from the **recording script**, excluding metadata and editorial notes. Count words, then report a range:

- calm conversational: 105–115 words/minute;
- expressive and measured: 90–95 words/minute;
- meditative: 80–85 words/minute;
- very slow trance: about 75 words/minute.

Then add realistic space for paragraph breaths, questions, Hacking Statements, music, and integration pauses. Give both `clean speech time` and `likely produced duration`; do not present one exact runtime before recording.

## 12. Terminology and positioning research

When Margarita asks whether a CL term is already used by other methods or markets:

1. search exact Russian and English forms plus natural synonyms;
2. separate **exact branded use**, **established professional term**, and **generic metaphor**;
3. inspect psychology, coaching, education, energetic/spiritual practice, leadership/HR, wellbeing, productivity, and AI-assisted development when relevant;
4. give direct URLs and describe the actual use rather than relying on snippets;
5. report a **confirmed minimum of distinct examples**, never an invented total for the whole world;
6. label the result as linguistic/market reconnaissance rather than trademark advice;
7. assess semantic crowding, search noise, likely audience confusion, and brand distinctiveness;
8. distinguish a decision for the current AudioBoost from a global change to the CL encyclopedia;
9. if replacing a system-wide metaphor such as `Human OS`, first map its occurrences and conceptual functions, then propose replacements that preserve the useful logic (settings, update, restart, resources, future integration) without blindly deleting the architecture.

For question terminology specifically, compare exact labels such as `создающие вопросы` and `энергетические вопросы` against Access-like syntax, generative coaching, NLP, Energy Psychology, education, and ordinary industry meanings. A crowded term may still be intentionally chosen for one AudioBoost; record that as local usage unless Margarita explicitly declares it canonical.

## 13. Versioning and verification

Once the approved batch is applied:

- update the internal version/status and task/description text yourself;
- update the preview title and description if the concept has evolved;
- verify the exact Hacking Statement count equals the number of cycles;
- search for rejected phrases and accidental old versions;
- verify question count per cycle;
- ensure the full script and extracted Hacking Process are synchronized;
- verify authorial opening, safe transition into relaxation, complete return, and signature ending;
- rebuild the preview and verify the delivered page contains the new wording.

Do not call a draft canonical merely because it has been placed in a preview. It becomes approved language only after Margarita’s authorial confirmation.