review-image-batch
Review a batch of images and output layout suggestions to an isolated batch file. Designed for parallel execution - each batch writes to its own file.
What this skill does
# Review Image Batch
Sub-agent skill for reviewing images and outputting layout blocks to an isolated file.
## Philosophy
Apply Szarkowski's Five Characteristics when assessing each image:
| Characteristic | Question to Ask |
|----------------|-----------------|
| **The Thing Itself** | What is the essential subject? What reality is captured? |
| **The Detail** | What fragments carry symbolic weight? |
| **The Frame** | What's included/excluded? How does framing create meaning? |
| **Time** | What moment is captured? Decisive or contemplative? |
| **Vantage Point** | How does position redistribute power/relationships? |
This vocabulary helps explain *why* an image deserves a particular layout.
## Key Design: Isolated Output
Each batch writes to its **own file** - no shared state, safe for parallel execution.
```
/tmp/gallery-batches/{folder}/
├── batch-00.json ← from batch 0
├── batch-01.json ← from batch 1
├── batch-02.json ← from batch 2
└── ...
```
## Usage
Invoked via Task tool from `/gallery` workflow:
```
Task(
subagent_type: "general-purpose",
prompt: "Review batch {N} for {folder} gallery..."
)
```
## Inputs
The sub-agent receives:
- `folder`: Gallery folder name
- `batchIndex`: Which batch (0, 1, 2, ...)
- `batchSize`: Images per batch (default 8)
## Workflow
1. **Read manifest** at `src/data/gallery-manifests/{folder}.json`
2. **Read layout guide** at `~/.claude/skills/review-image-batch/references/layout-guide.md`
3. **Calculate image range**: `startImage = batchIndex * batchSize`
4. **View images** using Read tool on each image file
5. **Create layout blocks** following the guide (80%+ single-image!)
6. **Write batch file** to `/tmp/gallery-batches/{folder}/batch-{NN}.json`
7. **Return text summary** only (no image data in response)
## Output File Format
```json
{
"batchIndex": 0,
"startImage": 0,
"endImage": 7,
"blocks": [
{
"layout": "FullBleed",
"images": ["DSCF9023.jpeg"],
"props": { "priority": true },
"notes": "Hero opener - kid with binoculars",
"visualWeight": "high"
},
{
"layout": "Spacer",
"images": [],
"props": { "size": "small" },
"notes": "Breathing room after hero",
"visualWeight": "rest"
},
{
"layout": "WideImage",
"images": ["DSCF9029.jpeg"],
"props": {},
"notes": "Joyful portrait - strong frame, decisive moment",
"visualWeight": "medium"
}
]
}
```
## Layout Types
| Layout | Images | When to Use |
|--------|--------|-------------|
| **FullBleed** | 1 | Hero shots, openers (max 2-4 per gallery) |
| **WideImage** | 1 | Strong standalone ← **DEFAULT CHOICE** |
| **OffsetImage** | 1 | Verticals, quiet moments |
| **InsetImage** | 1 | Maximum restraint, intimate |
| **Spacer** | 0 | Visual pause, breathing room |
| **Chapter** | 0 | Start new page section |
**RARE (use sparingly across entire gallery):**
| Layout | Images | Limit |
|--------|--------|-------|
| TwoUp | 2 | MAX 2-3 per gallery |
| ThreeUp | 3 | MAX 1-2 per gallery |
| SplitLayout | 2 | MAX 1-2 per gallery |
| FourUp | 4 | MAX 0-1 per gallery |
## Visual Weight Tracking
Track the weight of each block to ensure rhythm:
| Weight | Layouts |
|--------|---------|
| `high` | FullBleed |
| `medium` | WideImage, TwoUp, ThreeUp, SplitLayout |
| `low` | OffsetImage, InsetImage |
| `rest` | Spacer |
**Rule:** Never place consecutive `high` or `medium` blocks without a `low` or `rest` between.
## Critical Rules
1. **Default to single-image layouts** - 80%+ should be WideImage/OffsetImage/FullBleed
2. **Multi-image layouts are RARE** - only with EXTRAORDINARY reason
3. **"Related" is NOT enough** - images must lose meaning if separated
4. **Vertical images → OffsetImage**, not forced into pairs
5. **After FullBleed, add Spacer** for breathing room
6. **Look for Chapter breaks** - location/time/theme changes
### Lyons's Juxtaposition Test for TwoUp
Only pair when juxtaposition transforms meaning:
- Truly simultaneous moment?
- One frames the other?
- Formal echo (lines, shapes, colors)?
- Emotional contrast/reinforcement?
If unsure, **don't pair**.
## Block Schema
**Required fields:**
| Field | Type | Description |
|-------|------|-------------|
| `layout` | string | Component name (FullBleed, WideImage, etc.) |
| `images` | string[] | Filenames (not indices!) |
| `props` | object | Component props (can be `{}`) |
| `notes` | string | Brief description using Szarkowski vocabulary |
**Optional fields:**
| Field | Type | Description |
|-------|------|-------------|
| `visualWeight` | string | `high`, `medium`, `low`, `rest` |
## Example Response
```
Batch 2 complete. Reviewed images 16-23.
Created 7 blocks:
- WideImage: Street scene with pink building (strong frame, environmental)
- OffsetImage(left): Vertical portrait of vendor (intimate vantage)
- WideImage: Market stall with fruit (rich detail)
- Spacer(medium): Transition pause
- Chapter: "Uxmal" - location change detected (timestamp gap: 3 hours)
- FullBleed: Pyramid establishing shot (decisive moment, low vantage)
- WideImage: Carved stone detail (the detail characteristic)
Visual rhythm: medium → low → medium → rest → (break) → high → medium
Wrote to /tmp/gallery-batches/merida-2026/batch-02.json
```
## Composability
This skill is atomic and reusable:
- Used by `/gallery` for batch review phase
- Can review any set of images independently
- Output format works with `merge-gallery-batches.ts` script
## Reference
For detailed layout guidance, pairing rules, and rhythm patterns:
→ Read `references/layout-guide.md`
Related in Image & Video
watch
IncludedWatch a video (URL or local path). Downloads with yt-dlp, extracts auto-scaled frames with ffmpeg, pulls the transcript from captions (or Whisper API fallback), and hands the result to Claude so it can answer questions about what's in the video.
physical-ai-defect-image-generation
IncludedUse when the user wants to orchestrate defect image generation, run associated setup, or handle outputs on OSMO. The Day 0 path handles cold-start with USD-to-ROI, image-edit augmentation, and AnomalyGen to create initial PCBA datasets. The Day 1 path performs inference and labeling on real images. This skill helps with first-time asset setup, creation of finetuning checkpoints, and configuring deployment. Trigger keywords: defect image generation, dig workflow, dig pipeline, defect image detection workflow, aoi pipeline, aoi anomalygen, usd2roi anomalygen, day 0 pcba, day 1 pcba, day 1 real-photo alignment, day 1 manual roi, metal surface anomaly, glass defect, anomalygen finetune, setup_pcb, setup_metal, setup_glass, setup_pretrained, dig setup, dig datasets, dig pretrained checkpoint, dig image-edit endpoint.
accelint-react-best-practices
IncludedReact performance optimization and best practices. ALWAYS use this skill when working with any React code - writing components, hooks, JSX; refactoring; optimizing re-renders, memoization, state management; reviewing for performance; fixing hydration mismatches; debugging infinite re-renders, stale closures, input focus loss, animations restarting; preventing remounting; implementing transitions, lazy initialization, effect dependencies. Even simple React tasks benefit from these patterns. Covers React 19+ (useEffectEvent, Activity, ref props). Triggers - useEffect, useState, useMemo, useCallback, memo, inline components, nested components, components inside components, re-render, performance, hydration, SSR, Next.js, useDeferredValue, combined hooks.
elevenlabs-agents
IncludedBuild conversational AI voice agents with ElevenLabs Platform using React, JavaScript, React Native, or Swift SDKs. Configure agents, tools (client/server/MCP), RAG knowledge bases, multi-voice, and Scribe real-time STT. Use when: building voice chat interfaces, implementing AI phone agents with Twilio, configuring agent workflows or tools, adding RAG knowledge bases, testing with CLI "agents as code", or troubleshooting deprecated @11labs packages, Android audio cutoff, CSP violations, dynamic variables, or WebRTC config. Keywords: ElevenLabs Agents, ElevenLabs voice agents, AI voice agents, conversational AI, @elevenlabs/react, @elevenlabs/client, @elevenlabs/react-native, @elevenlabs/elevenlabs-js, @elevenlabs/agents-cli, elevenlabs SDK, voice AI, TTS, text-to-speech, ASR, speech recognition, turn-taking model, WebRTC voice, WebSocket voice, ElevenLabs conversation, agent system prompt, agent tools, agent knowledge base, RAG voice agents, multi-voice agents, pronunciation dictionary, voice speed control, elevenlabs scribe, @11labs deprecated, Android audio cutoff, CSP violation elevenlabs, dynamic variables elevenlabs, case-sensitive tool names, webhook authentication
humanizer
IncludedHumanize AI-generated text by detecting and removing patterns typical of LLM output. Rewrites text to sound natural, specific, and human. Uses 28 pattern detectors, 560+ AI vocabulary terms across 3 tiers, and statistical analysis (burstiness, type-token ratio, readability) for comprehensive detection. Use when asked to humanize text, de-AI writing, make content sound more natural/human, review writing for AI patterns, score text for AI detection, or improve AI-generated drafts. Covers content, language, style, communication, and filler categories.
generating-mermaid-diagrams
IncludedSalesforce architecture diagrams using Mermaid with ASCII fallback. Use this skill when generating text-based diagrams for Salesforce architecture, OAuth flows, ERDs, integration sequences, or Agentforce structure. TRIGGER when: user says "diagram", "visualize", "ERD", or asks for sequence diagrams, flowcharts, class diagrams, or architecture visualizations in Mermaid. DO NOT TRIGGER when: user wants PNG/SVG image output (use generating-visual-diagrams), or asks about non-Salesforce systems.