code-ai-installer 1.5.6 → 2.1.0

package/domains/content/locales/en/.agents/skills/board/SKILL.md

@@ -0,0 +1,195 @@
+---
+name: board
+description: Managing the content pipeline task board — initialization, statuses, event log, blocker escalation
+---
+# Board — Task Board Management
+
+## When to use
+- At pipeline **initialization** — Conductor creates the board as the first action.
+- At **each transition** between gates — in conjunction with `$handoff`.
+- On **blocker** — escalation to the user via the escalation protocol.
+- On **status request** — when the user asks "where are we now?".
+
+## Input
+
+| Field | Required | Description |
+|------|:-----------:|----------|
+| Pipeline mode | ✅ | Full / Bugfix / Hotfix |
+| Task name | ✅ | Brief name (feature, bug, content campaign) |
+| Task ID | ⬚ | External ID (Linear, Notion, etc.), if available |
+| Assignee | ⬚ | Who initiated the task |
+
+> If the mode is not determined — Conductor determines it by the Decision Tree prior to creating the board.
+
+## Statuses and State Machine
+
+### Allowed statuses
+
+| Icon | Status | Meaning |
+|--------|--------|----------|
+| `[ ]` | Pending | Gate has not started yet |
+| `[→]` | In progress | Agent is actively working on this gate |
+| `[✓]` | Completed | Gate passed, deliverable finalized, user sign-off received |
+| `[!]` | Blocker | Work halted, escalation required |
+| `[↩]` | Return | Gate sent back for revision (after review or user feedback) |
+
+### Allowed transitions
+
+```
+[ ] Pending ──→ [→] In progress      (agent started work)
+[→] In progress ──→ [✓] Completed     (deliverable ready + user "Approved")
+[→] In progress ──→ [!] Blocker       (blocker found)
+[→] In progress ──→ [↩] Return        (user requested revision)
+[!] Blocker ────→ [→] In progress      (blocker resolved)
+[↩] Return ───→ [→] In progress      (revision started)
+[✓] Completed ──→ [↩] Return          (late feedback, review)
+```
+
+> **Forbidden transitions:**
+> - `[ ] Pending` → `[✓] Completed` (skipping a gate is not allowed)
+> - `[!] Blocker` → `[✓] Completed` (blocker must be resolved first)
+> - Any backward transition across multiple gates (return — only to the current or previous)
+
+## Protocol
+
+### Step 1 — Board Initialization
+1. Obtain input data (mode, name, ID).
+2. Determine the set of gates based on the mode (see templates below).
+3. Create a board with all gates in status `[ ] Pending`.
+4. Transition the first gate (Conductor) to `[→] In progress`.
+5. Show the board to the user.
+
+### Step 2 — Update on transition
+On each `$handoff`:
+1. Current gate → `[✓] Completed`, record deliverable and date.
+2. Next gate → `[→] In progress`.
+3. Add a record to the event log.
+4. Show the updated board to the user.
+
+### Step 3 — Blocker handling
+1. Current gate → `[!] Blocker`.
+2. Add a description of the blocker in the event log.
+3. Execute the escalation protocol (see below).
+4. Upon resolution: gate → `[→] In progress`, add a record to the log.
+
+### Step 4 — Return handling
+1. Gate to which it returns → `[↩] Return`.
+2. All subsequent gates → `[ ] Pending` (reset).
+3. Document the reason for return in the event log.
+4. After revision: gate → `[→] In progress`.
+
+### Step 5 — Closing the board
+1. Ensure all gates are in `[✓] Completed`.
+2. Add a final record to the log.
+3. Show the final board + complete log.
+
+## Escalation Protocol
+
+Upon changing a gate to `[!] Blocker` — form a message:
+
+```
+### ⚠️ Blocker — [Gate]
+
+**Task:** [Task name]
+**Gate:** [Gate name]
+**Blocker description:** [What exactly is blocked and why]
+**Impact:** [Which subsequent gates it affects]
+**Solution options:**
+1. [Option A] — [pros / cons]
+2. [Option B] — [pros / cons]
+
+→ Awaiting decision from the user.
+```
+
+## Integration with other skills
+
+| Skill | Interaction |
+|------|----------------|
+| `$handoff` | On every handoff — update the board (step 2) |
+| `$gates` | Board reflects the gates defined in `$gates` |
+| `$content-release-gate` | Final gate of the board — Release Gate |
+
+## Validation (Quality Gate)
+
+The board is considered correct if:
+
+- [ ] Support for gates corresponds to the selected pipeline mode
+- [ ] Exactly one gate is in status `[→] In progress` (or zero, if all completed / blocker present)
+- [ ] No skipped gates (`[✓]` go sequentially from the beginning)
+- [ ] All `[✓]` gates have a populated deliverable and date
+- [ ] Event log contains a record for each transition
+- [ ] No forbidden transitions
+
+## Anti-patterns
+
+| Error | Why it's bad | How to do it right |
+|--------|-------------|---------------|
+| Skipping board update | Tracking is lost, unclear where we are | Update upon every `$handoff` |
+| Multiple `[→]` gates concurrently | Parallel work without control | Only one `[→]` at any given time |
+| `[✓]` without deliverable | Unclear what was done | Always fill in the "Deliverable" column |
+| Blocker without escalation | User is unaware of the problem | Always use the escalation protocol |
+| Board not shown to the user | User is out of context | Show the board upon each update |
+| Manual editing of past statuses | Distorting history | Only `[↩] Return`, no overwriting |
+
+## Output Templates
+
+### Full Pipeline
+
+```
+### Task Board — Full Pipeline
+**Task:** [Name]
+**ID:** [ID or —]
+
+| # | Gate | Status | Deliverable | Date |
+|---|------|--------|-------------|------|
+| 1 | Conductor | [→] In progress | — | [date] |
+| 2 | Strategist | [ ] Pending | — | — |
+| 3 | Researcher | [ ] Pending | — | — |
+| 4 | Copywriter | [ ] Pending | — | — |
+| 5 | Visual Concept | [ ] Pending | — | — |
+| 6 | Reviewer | [ ] Pending | — | — |
+| 7 | Release Gate | [ ] Pending | — | — |
+
+#### Event Log
+| Time | Gate | Event | Details |
+|-------|------|---------|--------|
+| [timestamp] | Conductor | [ ] → [→] | Pipeline initialized |
+```
+
+### Bugfix Pipeline
+
+```
+### Task Board — Bugfix Pipeline
+**Task:** [Name]
+**ID:** [ID or —]
+
+| # | Gate | Status | Deliverable | Date |
+|---|------|--------|-------------|------|
+| 1 | Conductor | [→] In progress | — | [date] |
+| 2 | Researcher | [ ] Pending | — | — |
+| 3 | Copywriter | [ ] Pending | — | — |
+| 4 | Reviewer | [ ] Pending | — | — |
+
+#### Event Log
+| Time | Gate | Event | Details |
+|-------|------|---------|--------|
+| [timestamp] | Conductor | [ ] → [→] | Pipeline initialized |
+```
+
+### Hotfix Pipeline
+
+```
+### Task Board — Hotfix Pipeline
+**Task:** [Name]
+**ID:** [ID or —]
+
+| # | Gate | Status | Deliverable | Date |
+|---|------|--------|-------------|------|
+| 1 | Conductor | [→] In progress | — | [date] |
+| 2 | Copywriter + Reviewer | [ ] Pending | — | — |
+
+#### Event Log
+| Time | Gate | Event | Details |
+|-------|------|---------|--------|
+| [timestamp] | Conductor | [ ] → [→] | Pipeline initialized |
+```

package/domains/content/locales/en/.agents/skills/brand-compliance/SKILL.md

@@ -0,0 +1,199 @@
+---
+name: brand-compliance
+description: Verification of content compliance with the brand — ToV, visual identity, messaging, legal — with severity levels and fix recommendations
+---
+# Brand Compliance
+
+## When to use
+- Upon **content review** prior to publication (Reviewer gate).
+- Upon **visuals check** — banners, creatives, stories, video covers.
+- Within **audit** of existing content for compliance with updated guidelines.
+- During **onboarding of a new copywriter / designer** — calibration of initial materials.
+
+## Input
+
+| Field | Required | Description |
+|------|:-----------:|----------|
+| Content for review | ✅ | Text, visual, or full post — what is being checked |
+| Brand guidelines | ✅ | Result of `$brand-guidelines` or brand guidelines file/URL |
+| ToV document | ✅ | Result of `$tone-of-voice` or description of the brand's tone |
+| Platform | ✅ | Where it will be published (affects format specifications) |
+| Content type | ⬚ | Post / article / email / banner / video / stories |
+| Previous remarks | ⬚ | If this is a re-check — the list of prior violations |
+
+> If brand guidelines or ToV are not provided — **request them from the user**. Verification is impossible without a benchmark.
+
+## Severity Levels
+
+| Level | Icon | Definition | Examples |
+|---------|--------|-------------|---------|
+| **Critical** | 🔴 | Publishing impossible. Legal risk, gross brand distortion, forbidden content. | Incorrect logo, legal disclaimer missing, offensive phrasing, copyright infringement |
+| **Major** | 🟡 | Publishing undesirable. Noticeable deviation from the brand, requires fixing. | Incorrect ToV, use of forbidden words, colors outside the palette, messaging mismatching positioning |
+| **Minor** | 🔵 | Publishing possible, but fixing is recommended. Cosmetic deviations. | Slightly different color shade, photo style differs slightly, minor stylistic discrepancies |
+
+### Passing Threshold
+
+| Result | Condition |
+|-----------|---------|
+| **COMPLIANT** | 0 Critical + 0 Major |
+| **CONDITIONALLY COMPLIANT** | 0 Critical + 1-2 Major (publishing after fixing Major) |
+| **NON-COMPLIANT** | ≥ 1 Critical OR ≥ 3 Major |
+
+## Protocol
+
+### Step 0 — Preparation
+1. Retrieve all mandatory input data.
+2. Load brand guidelines (`$brand-guidelines`) and ToV document (`$tone-of-voice`).
+3. If this is a re-check — load the list of prior violations.
+
+### Step 1 — Check Tone of Voice
+1. Compare text formality / informality against the ToV document.
+2. Check emotional coloring — does it align with the brand's character.
+3. Check for forbidden words, clichés, stop-phrases.
+4. Assess tone uniformity throughout the material.
+5. Check customer language — are words/phrases used that the TA understands (from `$audience-analysis`).
+
+### Step 2 — Check Visual Identity
+1. Color palette — all colors are within the approved palette (HEX codes).
+2. Typography — fonts, sizes, styles align with guidelines.
+3. Logo — size, clear space, permissible variant, forbidden transformations.
+4. Visual style — photos, illustrations, icons correspond to brand style.
+5. Composition — paddings, alignment, elements hierarchy.
+6. Platform specifications — sizes, aspect ratios, safe zones for the platform.
+
+### Step 3 — Check Messaging
+1. Value proposition — conveyed correctly, untampered.
+2. Positioning — aligns with brand strategy.
+3. CTA — lines up with approved phrasing, no brand discrepancies.
+4. Consistency — no conflicts with other brand content on the same channel.
+5. Key messages — priority brand theses are present (if applicable).
+
+### Step 4 — Check Legal and Compliance Requirements
+1. Legal disclaimer — present if required (advertising, finance, health).
+2. Copyright — © and year mentioned correctly.
+3. Disclaimer — if promises/guarantees are made.
+4. User-generated content — if external content used, is there permission.
+5. Brand hashtags — used approved ones, no forbidden ones.
+
+### Step 5 — Forming the report
+1. For each violation: severity + specific quote/example + recommendation to fix.
+2. Calculate the outcome via the passing threshold.
+3. Generate a summary and action list.
+
+## Validation (Quality Gate)
+
+Check is considered completed if:
+
+- [ ] All 4 layers are verified (ToV, Visual, Messaging, Legal)
+- [ ] Severity, quote, and suggested fix are specified for each violation
+- [ ] Outcome reflects the passing threshold rules (not subjective)
+- [ ] If `NON-COMPLIANT` or `CONDITIONALLY COMPLIANT` — an action list is generated
+- [ ] If a re-check — prior violations are verified for fixes
+
+## Handoff
+
+| Outcome | Routing |
+|------|---------|
+| **COMPLIANT** | → `$content-release-gate` (final gate before publishing) |
+| **CONDITIONALLY COMPLIANT** | → back to assignee (Copywriter / Visual Concept) with a specific list of Major fixes → re-check |
+| **NON-COMPLIANT** | → back to assignee with the complete list → update `$board` status to `[↩] Return` → re-check |
+
+When handing off — use `$handoff` with the full report attached.
+
+## Integration with other skills
+
+| Skill | Interaction |
+|------|----------------|
+| `$brand-guidelines` | **Benchmark source** — visual identity, logo, palette, typography |
+| `$tone-of-voice` | **Benchmark source** — tone, style, forbidden words |
+| `$audience-analysis` | Customer language — check that text speaks in the TA's language |
+| `$content-review-checklist` | Complementary skill — `$brand-compliance` checks the brand, `$content-review-checklist` checks content quality |
+| `$platform-compliance` | Complementary skill — `$brand-compliance` checks the brand, `$platform-compliance` checks platform rules |
+| `$board` | Update gate status by the check outcome |
+
+## Anti-patterns
+
+| Error | Why it's bad | How to do it right |
+|--------|-------------|---------------|
+| Verification without guidelines | Subjective opinion, not compliance | Always vet against specific benchmarks |
+| All violations equally leveled | Trivialities block publication | Apply severity: Critical / Major / Minor |
+| "A violation exists" without quote | Assignee clueless on what to fix | Always quote the specific place |
+| Violation without suggested fix | Feedback lacks resolution | Serve a recommendation per violation |
+| Text-only check | Visuals might breach the brand worse | Check all 4 layers |
+| Skipping legal | Penalties, bans, lawsuits | Always check disclaimer, copyright, permissions |
+
+## Output Template
+
+```
+### Brand Compliance — [Content / Publication]
+
+**Platform:** [platform]
+**Content type:** [type]
+**Benchmark:** [link to brand guidelines + ToV]
+**Re-check:** [Yes (iteration N) / No]
+
+---
+
+#### 1. Tone of Voice
+
+| Aspect | Requirement | Fact | Status |
+|--------|-----------|------|--------|
+| Formality | [level from ToV] | [observation + quote] | ✅/❌ |
+| Emotionality | [level from ToV] | [observation + quote] | ✅/❌ |
+| Forbidden words | None | [present/absent + quote] | ✅/❌ |
+| Tone uniformity | Consistent throughout | [observation] | ✅/❌ |
+| Customer language | [words from audience-analysis] | [observation] | ✅/❌ |
+
+#### 2. Visual Identity
+
+| Element | Requirement | Fact | Status |
+|---------|-----------|------|--------|
+| Color palette | [HEX-codes from guidelines] | [used colors] | ✅/❌ |
+| Typography | [font, size, style] | [used ones] | ✅/❌ |
+| Logo | [size, clear space, variant] | [fact] | ✅/❌ |
+| Visual style | [style for photos/art] | [fact] | ✅/❌ |
+| Composition | [paddings, hierarchy] | [fact] | ✅/❌ |
+| Platform specs | [sizes for platform] | [fact] | ✅/❌ |
+
+#### 3. Messaging
+
+| Aspect | Requirement | Fact | Status |
+|--------|-----------|------|--------|
+| Value proposition | [from strategy] | [observation] | ✅/❌ |
+| Positioning | [from strategy] | [observation] | ✅/❌ |
+| CTA | [approved phrasings] | [fact] | ✅/❌ |
+| Consistency | No conflicts | [observation] | ✅/❌ |
+
+#### 4. Legal and Compliance
+
+| Aspect | Required | Fact | Status |
+|--------|-----------|------|--------|
+| Disclaimer | [Yes/No] | [present/absent] | ✅/❌/N/A |
+| Copyright © | [Yes/No] | [present/absent] | ✅/❌/N/A |
+| Brand hashtags | [list] | [used ones] | ✅/❌/N/A |
+| UGC Permissions | [Yes/No] | [present/absent] | ✅/❌/N/A |
+
+---
+
+### Violations
+
+| # | Severity | Layer | Description | Quote / Example | Suggested Fix |
+|---|----------|------|----------|-----------------|---------------|
+| 1 | 🔴 Critical | Legal | [description] | "[quote]" | [how to fix] |
+| 2 | 🟡 Major | ToV | [description] | "[quote]" | [how to fix] |
+| 3 | 🔵 Minor | Visual | [description] | [example] | [how to fix] |
+
+**Summary:** 🔴 [N] Critical · 🟡 [N] Major · 🔵 [N] Minor
+
+---
+
+### Outcome: [ COMPLIANT / CONDITIONALLY COMPLIANT / NON-COMPLIANT ]
+
+**Justification:** [why such outcome — via the passing threshold rules]
+
+**Action list (if not COMPLIANT):**
+1. [Fix Critical #1 — ...]
+2. [Fix Major #2 — ...]
+
+**Next step:** [→ $content-release-gate / → back to Copywriter / → back to Visual Concept]
+```

package/domains/content/locales/en/.agents/skills/brand-guidelines/SKILL.md

@@ -0,0 +1,271 @@
+---
+name: brand-guidelines
+description: Complete brand guideline — identity, visual, ToV, messaging, legal, platform adaptations, Do/Don't
+---
+# Brand Guidelines
+
+## When to use
+- Upon **creating a brand from scratch** — forming the complete guideline.
+- Upon **onboarding** — handing over guidelines to a new copywriter, designer, contractor.
+- Upon **rebranding** — updating an existing guideline and outlining changes.
+- Upon **audit** — checking the completeness and relevance of the current guideline.
+- As a **benchmark** for `$brand-compliance` — without this document, compliance verification is impossible.
+
+## Input
+
+| Field | Required | Description |
+|------|:-----------:|----------|
+| Brand name | ✅ | Legal and communication name |
+| Niche / industry | ✅ | Industry in which the brand operates |
+| Existing materials | ✅ | Figma, PDF guideline, website, socials — any visual and style sources |
+| Mission / values | ⬚ | Provide if already formulated; if not, will be defined in Step 1 |
+| Target audience | ⬚ | Result of `$audience-analysis` or TA description |
+| Competitors | ⬚ | For positioning and differentiation |
+
+> If there are no existing materials (brand from scratch) — the agent builds the guideline via a user interview. Do not invent details without confirmation.
+
+## Data Sources
+1. **Existing brand assets** — website, social networks, presentations, packaging.
+2. **Figma / Sketch / PDF** — design system, if available.
+3. **User interview** — mission, values, positioning, tone (if not documented).
+4. **`$audience-analysis`** — who we speak to, in what language.
+5. **`$competitor-content-analysis`** — what we stand apart from.
+
+## Protocol
+
+### Step 1 — Brand Identity (Core)
+1. Record or formulate the **mission** (why the brand exists).
+2. Define **values** (3-5 key principles).
+3. Formulate **positioning** (for whom, what it does, how it differs).
+4. Identify the **brand archetype** (if applicable: Hero, Sage, Rebel, etc.).
+5. Document **key messages** (3-5 theses that the brand always broadcasts).
+
+### Step 2 — Naming & Verbal Identity
+1. Rules for writing the brand name (case, language, permissible abbreviations).
+2. Slogan / tagline (if any).
+3. Tone for product / service names.
+4. Transliteration rules (for multilingual brands).
+
+### Step 3 — Tone of Voice
+1. Define tone characteristics (4 axes: formality, emotionality, humor, expertise).
+2. For each axis — a scale from 1 to 5 with the brand's current position.
+3. Forbidden words and clichés (stop-list).
+4. Mandatory speech patterns (if any).
+5. Link to `$tone-of-voice` — this skill provides a detailed development of tone.
+
+> The ToV section in the guideline is a brief summary. Detailed elaboration is done via `$tone-of-voice`.
+
+### Step 4 — Visual Identity
+1. **Color palette** — primary, accent, neutral colors (HEX, RGB, CMYK).
+2. **Typography** — fonts for headings, body text, CTA, captions.
+3. **Logo** — variants, minimum sizes, clear space, forbidden transformations.
+4. **Visual style** — photography, illustrations, icons, patterns/textures.
+5. **Composition** — grid, paddings, element hierarchy.
+6. **Graphical devices** — signature frames, shapes, overlays (if any).
+
+### Step 5 — Do / Don't
+For every key element (logo, colors, tone, photos) — specific examples:
+- ✅ **Do:** proper usage with description.
+- ❌ **Don't:** improper usage with an explanation of why it is forbidden.
+
+> Minimum 2 Do's and 2 Don'ts for each element: logo, colors, typography, ToV.
+
+### Step 6 — Platform Adaptations
+For every priority platform (from `$platform-strategy`):
+- Sizes and aspect ratios for visuals.
+- Permissible logo variant.
+- Typography adaptations (mobile sizes, web-safe fonts).
+- Tone specifics (if the platform requires adaptation — LinkedIn is more formal than TikTok).
+
+### Step 7 — Legal and Trademark
+1. Trademark symbol (™ / ®) — when and where to use.
+2. Copyright — format, year, mandatory inclusion.
+3. Rules for third-party use (partners, media, influencers).
+4. Restrictions: what is strictly forbidden to do with brand elements.
+
+### Step 8 — Versioning
+1. Assign a guideline version.
+2. Record the date and author.
+3. If an update — describe what changed (changelog).
+
+## Validation (Quality Gate)
+
+The guideline is considered complete if:
+
+- [ ] Brand Identity is filled (mission, values, positioning, key messages)
+- [ ] Verbal Identity is recorded (name writing, slogan)
+- [ ] ToV is defined (4 axes + stop-list) or linked to `$tone-of-voice`
+- [ ] Visual identity is comprehensive (palette, typography, logo, style)
+- [ ] Do / Don't are present for: logo, colors, typography, ToV (minimum 2 each)
+- [ ] Platform adaptations exist for all priority platforms
+- [ ] Legal section is populated (trademark, copyright, third-party rules)
+- [ ] Version and date are recorded
+- [ ] Guideline is approved by the user
+
+## Handoff
+
+The result of this skill serves as the **benchmark** for:
+
+| Consumer | How it's used |
+|-------------|---------------|
+| `$brand-compliance` | Checks content against the guideline — **primary consumer** |
+| `$tone-of-voice` | Uses the ToV section as a starting point for detailed elaboration |
+| `$visual-brief` | Takes visual specifications for specific tasks |
+| `$content-brief` | Links to the guideline in the "Style" section |
+| Copywriter (agent) | Follows ToV and messaging when writing texts |
+| Visual Concept (agent) | Follows visual identity when creating creatives |
+
+When handing off — use `$handoff` specifying the guideline version.
+
+## Anti-patterns
+
+| Error | Why it's bad | How to do it right |
+|--------|-------------|---------------|
+| Guideline lacking mission/values | Visuals without a core — cosmetics without strategy | Always start with Brand Identity |
+| Visuals only, no text | Copywriter doesn't know how to write | ToV + messaging are mandatory |
+| Absent Do / Don't | Abstract rules — everyone misinterprets them | Specific examples of right and wrong |
+| One size fits all platforms | A 1000px logo in stories looks different than in a header | Adaptations for each platform |
+| No stop-list | Copywriter uses words foreign to the brand | Explicit list of forbidden words and clichés |
+| No versioning | Updates leave ambiguity over what is current | Always version + date + changelog |
+| Hallucinated guidelines | The agent fabricated a palette without the user | Confirm everything with the user |
+
+## Output Template
+
+```
+### Brand Guidelines — [Brand]
+**Version:** [N.N]
+**Date:** [YYYY-MM-DD]
+**Author:** [who created / updated]
+**Changelog:** [what changed in this version or "First version"]
+
+---
+
+#### 1. Brand Identity
+
+**Mission:** [why the brand exists — 1-2 sentences]
+
+**Values:**
+1. [Value 1] — [brief explanation]
+2. [Value 2] — [brief explanation]
+3. [Value 3] — [brief explanation]
+
+**Positioning:**
+> For [target audience], [brand] is a [category] that [key differentiator], because [reason to believe].
+
+**Archetype:** [archetype, if applicable]
+
+**Key messages:**
+1. [Thesis 1]
+2. [Thesis 2]
+3. [Thesis 3]
+
+---
+
+#### 2. Naming & Verbal Identity
+
+- **Writing:** [correct spelling of the name, case, variants]
+- **Slogan:** [tagline]
+- **Forbidden variants:** [how NOT to write the name]
+- **Transliteration:** [rules for other languages, if applicable]
+
+---
+
+#### 3. Tone of Voice (Summary)
+
+| Axis | 1 | 2 | 3 | 4 | 5 | Brand Position |
+|-----|---|---|---|---|---|----------------|
+| Formal ↔ Informal | | | | | | [●] on scale |
+| Serious ↔ Playful | | | | | | [●] on scale |
+| Restrained ↔ Emotional| | | | | | [●] on scale |
+| Practitioner ↔ Visionary| | | | | | [●] on scale |
+
+**Stop-list:** [forbidden words and clichés]
+**Mandatory patterns:** [signature phrases, if they exist]
+
+> Full tone elaboration → `$tone-of-voice`
+
+---
+
+#### 4. Visual Identity
+
+**Color Palette:**
+| Role | HEX | RGB | CMYK | Usage |
+|------|-----|-----|------|---------------|
+| Primary | #[hex] | [rgb] | [cmyk] | Headings, buttons |
+| Accent | #[hex] | [rgb] | [cmyk] | CTA, accents |
+| Neutral | #[hex] | [rgb] | [cmyk] | Background, text |
+| Sub-color | #[hex] | [rgb] | [cmyk] | Decorative elements |
+
+**Typography:**
+| Element | Font | Size | Weight | Fallback |
+|---------|-------|--------|------------|----------|
+| H1 | [font] | [px/rem] | [Bold] | [web-safe] |
+| H2 | [font] | [px/rem] | [SemiBold] | [web-safe] |
+| Body | [font] | [px/rem] | [Regular] | [web-safe] |
+| CTA | [font] | [px/rem] | [Bold] | [web-safe] |
+| Caption | [font] | [px/rem] | [Light] | [web-safe] |
+
+**Logo:**
+- Primary variant: [description]
+- Monochrome: [description]
+- Inverted: [description]
+- Favicon / icon: [description]
+- Minimum size: [N px]
+- Clear space: [N × logo element]
+- Forbidden: [list of transformations]
+
+**Visual Style:**
+- Photography: [style, filters, image types, mood]
+- Illustrations: [style, line thickness, fill]
+- Icons: [style, size, stroke/fill]
+- Patterns / textures: [description]
+- Graphical devices: [signature shapes, frames, overlays]
+
+---
+
+#### 5. Do / Don't
+
+**Logo:**
+| ✅ Do | ❌ Don't |
+|-------|---------|
+| [proper usage] | [improper usage] |
+| [proper usage] | [improper usage] |
+
+**Colors:**
+| ✅ Do | ❌ Don't |
+|-------|---------|
+| [proper usage] | [improper usage] |
+| [proper usage] | [improper usage] |
+
+**Typography:**
+| ✅ Do | ❌ Don't |
+|-------|---------|
+| [proper usage] | [improper usage] |
+| [proper usage] | [improper usage] |
+
+**Tone of Voice:**
+| ✅ Do | ❌ Don't |
+|-------|---------|
+| [example of correct tone + quote] | [example of incorrect tone + quote] |
+| [example of correct tone + quote] | [example of incorrect tone + quote] |
+
+---
+
+#### 6. Platform Adaptations
+
+| Platform | Logo (variant) | Visual sizes | Typography | Tone (adaptation) |
+|-----------|---------------|------------------|-------------|-----------------|
+| Instagram | [variant] | Feed: 1080×1080, Stories: 1080×1920, Reels: 1080×1920 | [adaptation] | [adaptation] |
+| Telegram | [variant] | Post: 800×400, Avatar: 640×640 | [adaptation] | [adaptation] |
+| LinkedIn | [variant] | Post: 1200×627, Banner: 1584×396 | [adaptation] | [adaptation] |
+| Web | [variant] | Header, OG:1200×630, Favicon:32×32 | [adaptation] | [adaptation] |
+
+---
+
+#### 7. Legal and Trademark
+
+- **Trademark:** [™ / ® — where and when to use]
+- **Copyright:** © [year] [legal entity]. Format: [description]
+- **Third parties:** [rules for influencers, partners, media]
+- **Strictly forbidden:** [list of absolute bans]
+```