wicked-garden 12.27.0

package/skills/product/imagery/review/refs/quality-gates.md

@@ -0,0 +1,153 @@
+---
+phase_relevance: ["clarify", "design", "review"]
+archetype_relevance: ["*"]
+---
+# Quality Gates Reference
+
+Four review gates ensure visual assets are "Business Ready" before delivery. Each gate produces a PASS/FAIL verdict with actionable feedback.
+
+---
+
+## Gate 1: Brand Compliance
+
+Ensures a visual asset aligns with corporate or project brand guidelines.
+
+### Checklist
+- [ ] **Logo placement**: Correct position, size, and clearance zone
+- [ ] **Logo integrity**: Not distorted, recolored, or cropped
+- [ ] **Primary brand colors**: Dominant and correctly applied
+- [ ] **Typography**: Correct fonts, weights, and sizes (if text present)
+- [ ] **Tone alignment**: Visual style matches brand archetype (e.g., luxury = refined lighting, tech = clean/modern)
+- [ ] **Photography style**: Matches brand guidelines (candid vs. staged, warm vs. cool)
+
+### Review Workflow
+1. **Palette Check:** Compare extracted hex codes against the official style guide
+2. **Logo Integrity:** Ensure the logo isn't distorted, recolored incorrectly, or placed too close to other elements
+3. **Aesthetic Alignment:** Does the image "feel" like the brand?
+
+### Pass Criteria
+- Primary brand colors are dominant
+- Visual style matches the established brand archetype
+- Imagery is appropriate for the target audience
+- No unauthorized modifications to brand elements
+
+### Common Failures
+| Issue | Fix |
+|-------|-----|
+| Off-brand color (#2196F3 vs brand #1976D2) | Adjust color values to match brand palette exactly |
+| Logo too small or crowded | Increase clearance zone, resize to minimum spec |
+| Wrong mood (casual for premium brand) | Adjust lighting, composition, and subject styling |
+
+---
+
+## Gate 2: Accessibility (a11y)
+
+Ensures visual assets are accessible to all users, including those with visual impairments.
+
+### Checklist
+- [ ] **Text contrast**: Meets WCAG 2.1 AA (4.5:1 for normal text, 3:1 for large text)
+- [ ] **Non-text contrast**: UI components and graphics meet 3:1 ratio
+- [ ] **Color independence**: Information not conveyed by color alone
+- [ ] **Color-blind safety**: Key elements distinguishable in protanopia/deuteranopia simulation
+- [ ] **Text readability**: Legible at intended display size
+- [ ] **Visual noise**: No distracting patterns or animations that could trigger seizures
+
+### Review Workflow
+1. **Contrast Analysis:** Check contrast between text/icons and their backgrounds
+2. **Color-Blind Simulation:** Analyze how the image appears in grayscale or simulated color-blindness modes
+3. **Readability Check:** Ensure any text within the image is legible at its intended display size
+
+### Pass Criteria
+- Contrast meets WCAG AA or AAA standards
+- Users can distinguish between elements without relying on color
+- Focal point is clear and undistorted
+- All text is legible at intended display size
+
+### WCAG Quick Reference
+| Element | AA Minimum | AAA Target |
+|---------|-----------|------------|
+| Normal text (< 18pt) | 4.5:1 | 7:1 |
+| Large text (≥ 18pt or 14pt bold) | 3:1 | 4.5:1 |
+| UI components / graphics | 3:1 | 3:1 |
+
+---
+
+## Gate 3: Production Readiness
+
+Final quality gate before a visual asset is delivered for business use.
+
+### Checklist
+- [ ] **Resolution**: Meets minimum spec for intended use (web: 72dpi, print: 300dpi)
+- [ ] **Aspect ratio**: Matches requested dimensions exactly
+- [ ] **File format**: Correct format (PNG for transparency, JPEG for photos, SVG for icons)
+- [ ] **File size**: Within acceptable limits for delivery channel
+- [ ] **Artifacts**: No blur, pixelation, or AI-generated distortions
+- [ ] **Cleanliness**: No unintended text, watermarks, or background clutter
+- [ ] **Edge quality**: Clean edges on subjects, no halo effects or fringing
+
+### Review Workflow
+1. **Artifact Inspection:** Zoom in on edges and complex textures to check for AI "hallucinations" or blurring
+2. **Spec Validation:** Confirm the image matches the requested aspect ratio and resolution
+3. **Background Cleanup:** Identify any distracting elements that don't serve the primary subject
+
+### Pass Criteria
+- Image is sharp and free of compression artifacts
+- Dimensions match the request exactly
+- Suitable for intended channel (social media, print, web, presentation)
+
+### Common Specifications
+| Channel | Resolution | Aspect Ratio | Format | Max Size |
+|---------|-----------|--------------|--------|----------|
+| Web hero | 1920x1080 | 16:9 | JPEG/WebP | 500KB |
+| Social (Instagram) | 1080x1080 | 1:1 | JPEG/PNG | 8MB |
+| Social (Story) | 1080x1920 | 9:16 | JPEG/PNG | 8MB |
+| Print A4 | 2480x3508 | ~1:1.41 | PDF/TIFF | - |
+| Presentation | 1920x1080 | 16:9 | PNG | 5MB |
+
+---
+
+## Gate 4: Cultural Sensitivity
+
+Ensures visual assets are appropriate for global audiences and free of unintended bias.
+
+### Checklist
+- [ ] **Symbols**: No symbols/gestures with unintended cultural meanings
+- [ ] **Representation**: Diverse and inclusive (where people are depicted)
+- [ ] **Stereotypes**: No reinforcement of cultural stereotypes
+- [ ] **Color meanings**: Colors appropriate for target markets (white = purity in West, mourning in parts of Asia)
+- [ ] **Religious/political**: No unintended religious or political symbolism
+- [ ] **Gender/age**: Balanced representation unless context-specific
+
+### Review Workflow
+1. **Symbolic Analysis:** Check for symbols, gestures, or icons with different meanings across cultures
+2. **Representation Audit:** Ensure people and settings reflect intended diversity without stereotypes
+3. **Global Context:** Consider color and symbol meanings in target markets
+
+### Pass Criteria
+- Image is free of culturally insensitive tropes
+- Aligns with global inclusivity standards
+- Symbols used correctly in their intended context
+- Representation is balanced and respectful
+
+### Regional Considerations
+| Element | Western | East Asian | Middle Eastern | Latin American |
+|---------|---------|-----------|----------------|----------------|
+| White | Purity, clean | Mourning | Purity | Peace |
+| Red | Danger, love | Luck, prosperity | Danger | Passion |
+| Thumbs up | Approval | Approval | Offensive (some regions) | Approval |
+| Hand gestures | Varies | Conservative | Very conservative | Expressive |
+
+---
+
+## Gate Selection Guide
+
+Not every asset needs all four gates:
+
+| Asset Type | Brand | a11y | Production | Sensitivity |
+|-----------|-------|------|------------|-------------|
+| Internal draft | - | - | ✓ | - |
+| UI component | ✓ | ✓ | ✓ | - |
+| External marketing | ✓ | ✓ | ✓ | ✓ |
+| Social media | ✓ | - | ✓ | ✓ |
+| Investor presentation | ✓ | - | ✓ | ✓ |
+| Documentation screenshot | - | ✓ | ✓ | - |

package/skills/product/imagery/scripts/analyze.sh

@@ -0,0 +1,23 @@
+#!/bin/bash
+# analyze.sh - Image analysis reads image files directly (multimodal vision)
+#
+# This script is intentionally minimal. The imagery/review sub-skill
+# performs image analysis by reading image files directly — no external
+# CLI or API is needed.
+#
+# For programmatic analysis via a provider, use:
+#   cstudio analyze image --image <path>
+
+IMAGE="$1"
+
+if [ -z "$IMAGE" ]; then
+  echo "Usage: $0 <image_path>"
+  echo ""
+  echo "Image analysis is performed by reading the image file directly."
+  echo "Use the imagery/review skill for multi-dimensional visual analysis."
+  exit 1
+fi
+
+echo "Image analysis for '$IMAGE' should be performed by reading the file directly."
+echo "See: skills/imagery/review/SKILL.md"
+exit 0

package/skills/product/imagery/scripts/edit.sh

@@ -0,0 +1,14 @@
+#!/bin/bash
+# edit.sh - Delegates to provider.py for multi-provider image editing
+PLUGIN_ROOT="${CLAUDE_PLUGIN_ROOT:-$(cd "$(dirname "$0")/../../.." && pwd)}"
+
+IMAGE="$1"
+PROMPT="$2"
+OUT_FILE="${3:-./output_edit.png}"
+
+if [ -z "$IMAGE" ] || [ -z "$PROMPT" ]; then
+  echo "Usage: $0 <source_image> <prompt> [output_file]"
+  exit 1
+fi
+
+python3 "${PLUGIN_ROOT}/skills/imagery/scripts/provider.py" edit --image "$IMAGE" --prompt "$PROMPT" --output "$OUT_FILE"

package/skills/product/imagery/scripts/generate.sh

@@ -0,0 +1,13 @@
+#!/bin/bash
+# generate.sh - Delegates to provider.py for multi-provider image generation
+PLUGIN_ROOT="${CLAUDE_PLUGIN_ROOT:-$(cd "$(dirname "$0")/../../.." && pwd)}"
+
+PROMPT="$1"
+OUT_FILE="${2:-./output.png}"
+
+if [ -z "$PROMPT" ]; then
+  echo "Usage: $0 <prompt> [output_file]"
+  exit 1
+fi
+
+python3 "${PLUGIN_ROOT}/skills/imagery/scripts/provider.py" generate --prompt "$PROMPT" --output "$OUT_FILE"