code-ai-installer 3.0.0 → 4.0.1-a

package/domains/content/locales/en/.agents/skills/visual-brief/SKILL.md

@@ -1,308 +1,323 @@
----
-name: visual-brief
-description: Visual Brief — detailed visual specs including composition, moodboard/brand alignment, AI prompt connections, A/B options, and accessibility
----
-# Visual Brief
-
-## When to Use
-- When creating **any visual content** — post, banner, Stories, cover, carousel, thumbnail.
-- As a **specification of the `$moodboard`** — the moodboard sets the style, the visual brief describes the specific task.
-- As **input for `$image-prompt-engineering`** — brief → prompt for AI generation.
-- As a **spec document for a designer** — a complete specification without the need for clarifying questions.
-
-## Input
-
-| Field | Required | Description |
-|-------|:--------:|-------------|
-| Content Brief | ✅ | From `$content-brief` — topic, key points, CTA, persona, platform |
-| Brand Guidelines | ✅ | From `$brand-guidelines` — palette, typography, logo, visual style |
-| Moodboard | ⬚ | From `$moodboard` — if created for a campaign/series (mood, references, extended palette) |
-| Platform & Format | ✅ | From `$platform-visual-specs` — dimensions, ratio, safe zones |
-| Text on Visual | ⬚ | If text will be on the image — the exact headline, CTA |
-| Executor | ⬚ | Designer / AI generation / Canva template — affects the level of detail |
-
-> If brand guidelines are missing — **request them**. A visual brief without a brand = a random visual.
-
-## Connection with other skills
-
-```
-$moodboard (campaign style)
-       ↓
-$visual-brief (Spec for specific visual) ← $content-brief (topic, CTA)
-       ↓                                 ← $brand-guidelines (palette, logo)
-       ↓                                 ← $platform-visual-specs (dimensions)
-       ↓
-$image-prompt-engineering (AI prompt)
-       ↓
-Visual Concept (agent) → generation / creation
-       ↓
-$content-review-checklist (#20-25) → check
-```
-
-## Protocol
-
-### Step 1 — Context and Goal
-1. Load `$content-brief` — understand the topic, persona, and CTA.
-2. Determine the **goal of the visual**:
-   - **Grab attention** — hook, stop the scroll.
-   - **Convey information** — infographic, scheme, data.
-   - **Evoke emotion** — mood, atmosphere, inspiration.
-   - **Amplify CTA** — button, arrow, focus on action.
-3. Define **what the viewer must understand in 1-3 seconds** (the main idea of the visual).
-
-### Step 2 — Central Subject
-1. Describe the **main element** of the visual — what is in focus.
-2. Describe **subject details** — specifically, not abstractly:
-   - ❌ "Person working"
-   - ✅ "Young woman in a navy blazer working on a laptop in a bright office, smiling"
-3. If there are multiple elements — describe the **hierarchy**: what is primary, what is secondary, what is background.
-
-### Step 3 — Composition
-Determine the arrangement of elements:
-
-| Composition Element | Description | When to use |
-|---------------------|-------------|-------------|
-| **Focal point** | Where the viewer's eye should land first | Always — define the main anchor |
-| **Rule of thirds** | Subject on the intersection of thirds | Dynamic, editorial photos |
-| **Central** | Subject in the center, symmetry | Minimalism, product, logo |
-| **Negative space** | Empty space for text or breathing room | If there is text on the image |
-| **Text placement** | Where text will be: top / bottom / left / right / overlay | If there is a headline/CTA on the visual |
-| **Z-pattern / F-pattern**| Eye movement across the visual | Complex visuals with multiple elements |
-| **Grid** | Carousel or series — uniform grid | Carousels, post series |
-
-### Step 4 — Brand Elements
-From `$brand-guidelines`:
-1. **Colors** — specific HEX codes. Roles: primary, accent, background.
-2. **Typography** — font, size, weight (for text on the visual).
-3. **Logo** — is it needed? Which variant? Position? Size?
-4. **Graphic elements** — signature frames, overlays, patterns (if present in guidelines).
-
-From `$moodboard` (if applicable):
-5. **Style** — medium (photo / illustration / 3D), treatment, textures.
-6. **Mood** — emotion derived from the moodboard.
-
-### Step 5 — Text on Image
-If the visual contains text:
-
-| Element | Text | Size | Position | Style |
-|---------|------|------|----------|-------|
-| **Headline** | [text or "none"] | [large / medium] | [where on visual] | [weight, color] |
-| **Subheadline** | [text or "none"] | [medium / small] | [where] | [style] |
-| **CTA** | [text or "none"] | [size] | [where] | [button / text / arrow] |
-| **Watermark / ©**| [text or "none"] | [small] | [corner] | [semi-transparent] |
-
-**Text Rules:**
-- Readability: contrast ≥ 4.5:1 (use a background shape/overlay if needed).
-- Size: ≥ 16px on mobile (≈ 24px+ on source file).
-- Word count: headline ≤ 7 words, subheadline ≤ 15 words.
-
-### Step 6 — Constraints (Do / Don't)
-
-| ✅ Mandatory | ❌ Forbidden |
-|--------------|--------------|
-| [what must be there] | [what is not allowed] |
-| [element / style] | [element / style — reason why] |
-
-### Step 7 — References
-1. From `$moodboard` — select the **2-3 most relevant** for this specific task.
-2. For each — **what exactly to take** (palette? composition? style?).
-3. If there is no moodboard — describe the style with words + provide 1-2 references.
-
-### Step 8 — Technical Specs
-From `$platform-visual-specs`:
-1. Dimensions (W×H px).
-2. Ratio.
-3. File format.
-4. Safe zones.
-5. For AI generation: Midjourney `--ar` parameter.
-
-### Step 9 — Variants (A/B)
-If testing visuals is required:
-1. **Variant A** — primary (based on brief).
-2. **Variant B** — alternative (different composition / color accent / different subject).
-3. Specify the **test variable**: exactly what is different.
-
-### Step 10 — Accessibility
-1. **Alt text** — description of the image for screen readers.
-2. **Contrast** — text on the image ≥ 4.5:1.
-3. **Dark mode** — how will the visual look on a dark background?
-4. **Not just color** — information is not conveyed by color alone (use icons, labels).
-
-## Validation (Quality Gate)
-
-The Visual Brief is considered complete if:
-
-- [ ] Visual goal is defined (what the viewer understands in 1-3 seconds)
-- [ ] Central subject is described specifically (not "person", but detailed)
-- [ ] Composition is defined (focal point, placement, negative space for text)
-- [ ] Brand elements from `$brand-guidelines` are included (HEX colors, font, logo — position and variant)
-- [ ] Text on image is defined: content + size + position (or explicitly "no text")
-- [ ] Tech specs from `$platform-visual-specs` are included (size, ratio, format, safe zone)
-- [ ] Do/Don't constraints are listed
-- [ ] References are provided (from `$moodboard` or descriptively)
-- [ ] Alt text is prepared
-- [ ] Text contrast is ≥ 4.5:1 (if text is on the visual)
-- [ ] If A/B — the test variable is specified
-
-## Handoff
-
-| Consumer | How it is used |
-|----------|----------------|
-| `$image-prompt-engineering` | **Primary:** brief → prompt for AI generation |
-| Visual Concept (agent) | Spec for creating the visual (designer or AI) |
-| `$content-review-checklist` | Reviewer checks the visual against points #20-25 |
-| `$content-release-gate` | Visual as part of the publication package |
-
-When transferring — use `$handoff` and attach the full brief.
-
-## Integration with other skills
-
-| Skill | Interaction |
-|-------|-------------|
-| `$content-brief` | **Source:** topic, CTA, persona → visual goal and content |
-| `$brand-guidelines` | **Source:** palette, typography, logo, visual style |
-| `$moodboard` | **Source:** mood, references, extended palette |
-| `$platform-visual-specs` | **Source:** dimensions, ratio, safe zones |
-| `$image-prompt-engineering`| **Consumer:** brief → prompt |
-| `$content-review-checklist` | **Consumer:** verifying the visual |
-| `$headline-formulas` | **Source:** headline for the text on the visual |
-
-## Anti-patterns
-
-| Mistake | Why it's bad | How to do it right |
-|---------|--------------|--------------------|
-| "A pretty picture about marketing"| Abstract — designer/AI won't know what to do | Concrete subject, composition, mood |
-| Brief missing brand elements| Visual will not be recognizable as the brand | Always: HEX colors, font, logo from `$brand-guidelines` |
-| Text without position | Text covers a face / key element | Text placement + negative space in composition |
-| No safe zones | Text hidden behind platform UI | Safe zones from `$platform-visual-specs` |
-| Only 1 variant, no alternative | No options — if it fails, start from scratch | A/B: min 2 variants for key visuals |
-| Brief for AI = brief for designer| Different levels of specificity are needed | For AI → use `$image-prompt-engineering` with technical parameters |
-| No alt text | Inaccessible to screen readers, worse SEO | Always prepare alt text |
-| Small text on visual | Unreadable on mobile | ≥ 24px on source file, verify on a phone screen |
-
-## Output Template
-
-```markdown
-### Visual Brief — [Asset Name]
-
-**Content Brief:** [ID / topic from $content-brief]
-**Platform:** [platform]
-**Dimensions:** [W × H px]
-**Ratio:** [ratio]
-**Format:** [JPG / PNG / WebP / MP4]
-**Safe zone:** [margins — from $platform-visual-specs]
-**Executor:** [designer / AI generation / template]
-**Moodboard:** [link to $moodboard or "none"]
-
----
-
-#### 1. Concept
-
-- **Visual Goal:** [attract / inform / emotion / amplify CTA]
-- **Main Idea (in 1-3 sec):** [what the viewer must understand]
-- **Mood:** [emotion from $moodboard or description]
-
----
-
-#### 2. Central Subject
-
-**Main Element:**
-[Detailed description of the subject — specifically, as if for a photographer or AI]
-
-**Hierarchy of Elements:**
-| Level | Element | Description |
-|-------|---------|-------------|
-| Primary | [subject] | [details] |
-| Secondary | [element] | [details] |
-| Background| [background]| [color / environment / texture] |
-
----
-
-#### 3. Composition
-
-| Parameter | Value |
-|-----------|-------|
-| Focal point | [where on visual — top-center / right third / ...] |
-| Technique | [rule of thirds / central / symmetry / ...] |
-| Negative space | [where to leave room for text / breathing room] |
-| Text placement | [where text will be — top / bottom / left / overlay] |
-| Eye movement | [from where → to where: headline → image → CTA] |
-
----
-
-#### 4. Brand Elements
-
-**Colors (from $brand-guidelines):**
-| Role | HEX | Where used |
-|------|-----|------------|
-| Primary | #[hex] | [elements] |
-| Accent | #[hex] | [elements] |
-| Background | #[hex] | [background / overlay] |
-
-**Typography (if text is on the visual):**
-| Element | Font | Size | Weight | Color |
-|---------|------|------|--------|-------|
-| Headline | [font] | [px] | [Bold] | [HEX] |
-| CTA | [font] | [px] | [SemiBold] | [HEX] |
-
-**Logo:**
-- Needed: [yes / no]
-- Variant: [primary / monochrome / inverted]
-- Position: [corner / size]
-
-**Graphic Elements:** [frames / overlays / patterns — from $brand-guidelines or $moodboard]
-
----
-
-#### 5. Text on Image
-
-| Element | Text | Size | Position | Style |
-|---------|------|------|----------|-------|
-| Headline | [text / "none"] | [large] | [top-center] | [Bold, #HEX] |
-| Subheadline | [text / "none"] | [medium] | [below headline] | [Regular, #HEX] |
-| CTA | [text / "none"] | [medium] | [bottom] | [button: bg #HEX, text #HEX] |
-
-**Contrast:** [verified ≥ 4.5:1 / background overlay needed]
-
----
-
-#### 6. Constraints
-
-| ✅ Mandatory | ❌ Forbidden |
-|--------------|--------------|
-| [element] | [element — reason] |
-| [element] | [element — reason] |
-
----
-
-#### 7. References
-
-| # | Description | From $moodboard? | What to borrow |
-|---|-------------|:----------------:|----------------|
-| 1 | [visual description] | ✅ / No | [palette / composition / style] |
-| 2 | [description] | ✅ / No | [what to borrow] |
-
----
-
-#### 8. Variants (if A/B testing)
-
-| Variant | Difference from primary | Test Variable |
-|---------|-------------------------|---------------|
-| A (Primary) | — | — |
-| B | [different composition / color / subject] | [what we are testing] |
-
----
-
-#### 9. Accessibility
-
-| Parameter | Value |
-|-----------|-------|
-| Alt text | [description for screen readers] |
-| Text contrast | [≥ 4.5:1 ✅ / overlay needed] |
-| Dark mode | [verified / adaptation needed] |
-| Not just color | [✅ information is duplicated with icons/labels] |
-
----
-
-**→ Next Step:** → `$image-prompt-engineering` (if AI) / → Designer (if manual) / → `$content-review-checklist` (after creation)
-```
+---
+name: visual-brief
+description: Visual Brief — detailed visual specs including composition, moodboard/brand alignment, AI prompt connections, A/B options, and accessibility
+type: triggered
+domain: content
+owners:
+  - visual_concept
+gates:
+  - VISUAL_CONCEPT
+tech: []
+topic: []
+triggers:
+  - "visual-brief"
+  - "Визуальный бриф"
+  - "Визуальный бриф — детальное ТЗ на визуал с композицией, привязкой к moodboard/brand, AI-промпт связкой, A/B вариантами, accessibility"
+related: []
+budget_lines: 323
+schema_version: 1
+---
+# Visual Brief
+
+## When to Use
+- When creating **any visual content** — post, banner, Stories, cover, carousel, thumbnail.
+- As a **specification of the `$moodboard`** — the moodboard sets the style, the visual brief describes the specific task.
+- As **input for `$image-prompt-engineering`** — brief → prompt for AI generation.
+- As a **spec document for a designer** — a complete specification without the need for clarifying questions.
+
+## Input
+
+| Field | Required | Description |
+|-------|:--------:|-------------|
+| Content Brief | ✅ | From `$content-brief` — topic, key points, CTA, persona, platform |
+| Brand Guidelines | ✅ | From `$brand-guidelines` — palette, typography, logo, visual style |
+| Moodboard | ⬚ | From `$moodboard` — if created for a campaign/series (mood, references, extended palette) |
+| Platform & Format | ✅ | From `$platform-visual-specs` — dimensions, ratio, safe zones |
+| Text on Visual | ⬚ | If text will be on the image — the exact headline, CTA |
+| Executor | ⬚ | Designer / AI generation / Canva template — affects the level of detail |
+
+> If brand guidelines are missing — **request them**. A visual brief without a brand = a random visual.
+
+## Connection with other skills
+
+```
+$moodboard (campaign style)
+       ↓
+$visual-brief (Spec for specific visual) ← $content-brief (topic, CTA)
+       ↓                                 ← $brand-guidelines (palette, logo)
+       ↓                                 ← $platform-visual-specs (dimensions)
+       ↓
+$image-prompt-engineering (AI prompt)
+       ↓
+Visual Concept (agent) → generation / creation
+       ↓
+$content-review-checklist (#20-25) → check
+```
+
+## Protocol
+
+### Step 1 — Context and Goal
+1. Load `$content-brief` — understand the topic, persona, and CTA.
+2. Determine the **goal of the visual**:
+   - **Grab attention** — hook, stop the scroll.
+   - **Convey information** — infographic, scheme, data.
+   - **Evoke emotion** — mood, atmosphere, inspiration.
+   - **Amplify CTA** — button, arrow, focus on action.
+3. Define **what the viewer must understand in 1-3 seconds** (the main idea of the visual).
+
+### Step 2 — Central Subject
+1. Describe the **main element** of the visual — what is in focus.
+2. Describe **subject details** — specifically, not abstractly:
+   - ❌ "Person working"
+   - ✅ "Young woman in a navy blazer working on a laptop in a bright office, smiling"
+3. If there are multiple elements — describe the **hierarchy**: what is primary, what is secondary, what is background.
+
+### Step 3 — Composition
+Determine the arrangement of elements:
+
+| Composition Element | Description | When to use |
+|---------------------|-------------|-------------|
+| **Focal point** | Where the viewer's eye should land first | Always — define the main anchor |
+| **Rule of thirds** | Subject on the intersection of thirds | Dynamic, editorial photos |
+| **Central** | Subject in the center, symmetry | Minimalism, product, logo |
+| **Negative space** | Empty space for text or breathing room | If there is text on the image |
+| **Text placement** | Where text will be: top / bottom / left / right / overlay | If there is a headline/CTA on the visual |
+| **Z-pattern / F-pattern**| Eye movement across the visual | Complex visuals with multiple elements |
+| **Grid** | Carousel or series — uniform grid | Carousels, post series |
+
+### Step 4 — Brand Elements
+From `$brand-guidelines`:
+1. **Colors** — specific HEX codes. Roles: primary, accent, background.
+2. **Typography** — font, size, weight (for text on the visual).
+3. **Logo** — is it needed? Which variant? Position? Size?
+4. **Graphic elements** — signature frames, overlays, patterns (if present in guidelines).
+
+From `$moodboard` (if applicable):
+5. **Style** — medium (photo / illustration / 3D), treatment, textures.
+6. **Mood** — emotion derived from the moodboard.
+
+### Step 5 — Text on Image
+If the visual contains text:
+
+| Element | Text | Size | Position | Style |
+|---------|------|------|----------|-------|
+| **Headline** | [text or "none"] | [large / medium] | [where on visual] | [weight, color] |
+| **Subheadline** | [text or "none"] | [medium / small] | [where] | [style] |
+| **CTA** | [text or "none"] | [size] | [where] | [button / text / arrow] |
+| **Watermark / ©**| [text or "none"] | [small] | [corner] | [semi-transparent] |
+
+**Text Rules:**
+- Readability: contrast ≥ 4.5:1 (use a background shape/overlay if needed).
+- Size: ≥ 16px on mobile (≈ 24px+ on source file).
+- Word count: headline ≤ 7 words, subheadline ≤ 15 words.
+
+### Step 6 — Constraints (Do / Don't)
+
+| ✅ Mandatory | ❌ Forbidden |
+|--------------|--------------|
+| [what must be there] | [what is not allowed] |
+| [element / style] | [element / style — reason why] |
+
+### Step 7 — References
+1. From `$moodboard` — select the **2-3 most relevant** for this specific task.
+2. For each — **what exactly to take** (palette? composition? style?).
+3. If there is no moodboard — describe the style with words + provide 1-2 references.
+
+### Step 8 — Technical Specs
+From `$platform-visual-specs`:
+1. Dimensions (W×H px).
+2. Ratio.
+3. File format.
+4. Safe zones.
+5. For AI generation: Midjourney `--ar` parameter.
+
+### Step 9 — Variants (A/B)
+If testing visuals is required:
+1. **Variant A** — primary (based on brief).
+2. **Variant B** — alternative (different composition / color accent / different subject).
+3. Specify the **test variable**: exactly what is different.
+
+### Step 10 — Accessibility
+1. **Alt text** — description of the image for screen readers.
+2. **Contrast** — text on the image ≥ 4.5:1.
+3. **Dark mode** — how will the visual look on a dark background?
+4. **Not just color** — information is not conveyed by color alone (use icons, labels).
+
+## Validation (Quality Gate)
+
+The Visual Brief is considered complete if:
+
+- [ ] Visual goal is defined (what the viewer understands in 1-3 seconds)
+- [ ] Central subject is described specifically (not "person", but detailed)
+- [ ] Composition is defined (focal point, placement, negative space for text)
+- [ ] Brand elements from `$brand-guidelines` are included (HEX colors, font, logo — position and variant)
+- [ ] Text on image is defined: content + size + position (or explicitly "no text")
+- [ ] Tech specs from `$platform-visual-specs` are included (size, ratio, format, safe zone)
+- [ ] Do/Don't constraints are listed
+- [ ] References are provided (from `$moodboard` or descriptively)
+- [ ] Alt text is prepared
+- [ ] Text contrast is ≥ 4.5:1 (if text is on the visual)
+- [ ] If A/B — the test variable is specified
+
+## Handoff
+
+| Consumer | How it is used |
+|----------|----------------|
+| `$image-prompt-engineering` | **Primary:** brief → prompt for AI generation |
+| Visual Concept (agent) | Spec for creating the visual (designer or AI) |
+| `$content-review-checklist` | Reviewer checks the visual against points #20-25 |
+| `$content-release-gate` | Visual as part of the publication package |
+
+When transferring — use `$handoff` and attach the full brief.
+
+## Integration with other skills
+
+| Skill | Interaction |
+|-------|-------------|
+| `$content-brief` | **Source:** topic, CTA, persona → visual goal and content |
+| `$brand-guidelines` | **Source:** palette, typography, logo, visual style |
+| `$moodboard` | **Source:** mood, references, extended palette |
+| `$platform-visual-specs` | **Source:** dimensions, ratio, safe zones |
+| `$image-prompt-engineering`| **Consumer:** brief → prompt |
+| `$content-review-checklist` | **Consumer:** verifying the visual |
+| `$headline-formulas` | **Source:** headline for the text on the visual |
+
+## Anti-patterns
+
+| Mistake | Why it's bad | How to do it right |
+|---------|--------------|--------------------|
+| "A pretty picture about marketing"| Abstract — designer/AI won't know what to do | Concrete subject, composition, mood |
+| Brief missing brand elements| Visual will not be recognizable as the brand | Always: HEX colors, font, logo from `$brand-guidelines` |
+| Text without position | Text covers a face / key element | Text placement + negative space in composition |
+| No safe zones | Text hidden behind platform UI | Safe zones from `$platform-visual-specs` |
+| Only 1 variant, no alternative | No options — if it fails, start from scratch | A/B: min 2 variants for key visuals |
+| Brief for AI = brief for designer| Different levels of specificity are needed | For AI → use `$image-prompt-engineering` with technical parameters |
+| No alt text | Inaccessible to screen readers, worse SEO | Always prepare alt text |
+| Small text on visual | Unreadable on mobile | ≥ 24px on source file, verify on a phone screen |
+
+## Output Template
+
+```markdown
+### Visual Brief — [Asset Name]
+
+**Content Brief:** [ID / topic from $content-brief]
+**Platform:** [platform]
+**Dimensions:** [W × H px]
+**Ratio:** [ratio]
+**Format:** [JPG / PNG / WebP / MP4]
+**Safe zone:** [margins — from $platform-visual-specs]
+**Executor:** [designer / AI generation / template]
+**Moodboard:** [link to $moodboard or "none"]
+
+---
+
+#### 1. Concept
+
+- **Visual Goal:** [attract / inform / emotion / amplify CTA]
+- **Main Idea (in 1-3 sec):** [what the viewer must understand]
+- **Mood:** [emotion from $moodboard or description]
+
+---
+
+#### 2. Central Subject
+
+**Main Element:**
+[Detailed description of the subject — specifically, as if for a photographer or AI]
+
+**Hierarchy of Elements:**
+| Level | Element | Description |
+|-------|---------|-------------|
+| Primary | [subject] | [details] |
+| Secondary | [element] | [details] |
+| Background| [background]| [color / environment / texture] |
+
+---
+
+#### 3. Composition
+
+| Parameter | Value |
+|-----------|-------|
+| Focal point | [where on visual — top-center / right third / ...] |
+| Technique | [rule of thirds / central / symmetry / ...] |
+| Negative space | [where to leave room for text / breathing room] |
+| Text placement | [where text will be — top / bottom / left / overlay] |
+| Eye movement | [from where → to where: headline → image → CTA] |
+
+---
+
+#### 4. Brand Elements
+
+**Colors (from $brand-guidelines):**
+| Role | HEX | Where used |
+|------|-----|------------|
+| Primary | #[hex] | [elements] |
+| Accent | #[hex] | [elements] |
+| Background | #[hex] | [background / overlay] |
+
+**Typography (if text is on the visual):**
+| Element | Font | Size | Weight | Color |
+|---------|------|------|--------|-------|
+| Headline | [font] | [px] | [Bold] | [HEX] |
+| CTA | [font] | [px] | [SemiBold] | [HEX] |
+
+**Logo:**
+- Needed: [yes / no]
+- Variant: [primary / monochrome / inverted]
+- Position: [corner / size]
+
+**Graphic Elements:** [frames / overlays / patterns — from $brand-guidelines or $moodboard]
+
+---
+
+#### 5. Text on Image
+
+| Element | Text | Size | Position | Style |
+|---------|------|------|----------|-------|
+| Headline | [text / "none"] | [large] | [top-center] | [Bold, #HEX] |
+| Subheadline | [text / "none"] | [medium] | [below headline] | [Regular, #HEX] |
+| CTA | [text / "none"] | [medium] | [bottom] | [button: bg #HEX, text #HEX] |
+
+**Contrast:** [verified ≥ 4.5:1 / background overlay needed]
+
+---
+
+#### 6. Constraints
+
+| ✅ Mandatory | ❌ Forbidden |
+|--------------|--------------|
+| [element] | [element — reason] |
+| [element] | [element — reason] |
+
+---
+
+#### 7. References
+
+| # | Description | From $moodboard? | What to borrow |
+|---|-------------|:----------------:|----------------|
+| 1 | [visual description] | ✅ / No | [palette / composition / style] |
+| 2 | [description] | ✅ / No | [what to borrow] |
+
+---
+
+#### 8. Variants (if A/B testing)
+
+| Variant | Difference from primary | Test Variable |
+|---------|-------------------------|---------------|
+| A (Primary) | — | — |
+| B | [different composition / color / subject] | [what we are testing] |
+
+---
+
+#### 9. Accessibility
+
+| Parameter | Value |
+|-----------|-------|
+| Alt text | [description for screen readers] |
+| Text contrast | [≥ 4.5:1 ✅ / overlay needed] |
+| Dark mode | [verified / adaptation needed] |
+| Not just color | [✅ information is duplicated with icons/labels] |
+
+---
+
+**→ Next Step:** → `$image-prompt-engineering` (if AI) / → Designer (if manual) / → `$content-review-checklist` (after creation)
+```