Skip to content

Universal SEO skill for Claude Code. 26 sub-skills + 19 sub-agents covering technical SEO, E-E-A-T, schema, GEO/AEO, agent readiness (Lighthouse Agentic Browsing, WebMCP, llms.txt), backlinks, local SEO, e-commerce, international SEO, Google APIs, and PDF/Excel reporting. 9 optional extensions for live SEO data.

feat: add banana image generation extension (seo-image-gen)

Daniel36aac5ceeefd
diff --git a/.claude-plugin/plugin.json b/.claude-plugin/plugin.json
index 392f946..aedcf3e 100644
--- a/.claude-plugin/plugin.json
+++ b/.claude-plugin/plugin.json
@@ -1,6 +1,6 @@
{
"name": "claude-seo",
- "version": "1.4.0",
+ "version": "1.5.0",
"description": "Comprehensive SEO analysis skill for Claude Code. Covers technical SEO, content quality (E-E-A-T), schema markup, image optimization, sitemap architecture, AI search optimization (GEO), and strategic planning across all industries.",
"author": "AgriciDaniel",
"license": "MIT",
@@ -19,7 +19,9 @@
"skills/seo-plan/SKILL.md",
"skills/seo-programmatic/SKILL.md",
"skills/seo-competitor-pages/SKILL.md",
- "skills/seo-hreflang/SKILL.md"
+ "skills/seo-hreflang/SKILL.md",
+ "extensions/dataforseo/skills/seo-dataforseo/SKILL.md",
+ "extensions/banana/skills/seo-image-gen/SKILL.md"
],
"agents": [
"agents/seo-technical.md",
@@ -28,6 +30,8 @@
"agents/seo-sitemap.md",
"agents/seo-performance.md",
"agents/seo-visual.md",
- "agents/seo-geo.md"
+ "agents/seo-geo.md",
+ "extensions/dataforseo/agents/seo-dataforseo.md",
+ "extensions/banana/agents/seo-image-gen.md"
]
}
diff --git a/CLAUDE.md b/CLAUDE.md
index 0e6c9ea..bc9c25a 100644
--- a/CLAUDE.md
+++ b/CLAUDE.md
@@ -13,7 +13,7 @@ schema markup, image optimization, sitemap architecture, and AI search optimizat
```
claude-seo/
CLAUDE.md # Project instructions (this file)
- .claude-plugin/plugin.json # Plugin manifest (v1.3.2)
+ .claude-plugin/plugin.json # Plugin manifest v1.5.0
seo/ # Main orchestrator skill
SKILL.md # Entry point, routing table, core rules
references/ # On-demand knowledge files
@@ -41,6 +41,9 @@ claude-seo/
seo-performance.md # Core Web Vitals, page speed
seo-visual.md # Screenshots, mobile rendering
seo-geo.md # AI crawler access, GEO, citability
+ extensions/ # Optional add-on capabilities
+ dataforseo/ # Live SEO data via DataForSEO MCP
+ banana/ # AI image generation via Gemini MCP
docs/ # Extended documentation
ARCHITECTURE.md # System design overview
COMMANDS.md # Full command reference
@@ -65,6 +68,7 @@ claude-seo/
| `/seo programmatic` | Programmatic SEO analysis and planning |
| `/seo competitor-pages` | Competitor comparison page generation |
| `/seo hreflang <url>` | International SEO / hreflang audit |
+| `/seo image-gen [use-case] <desc>` | AI image generation for SEO assets (extension) |
## Development Rules
@@ -81,4 +85,4 @@ claude-seo/
1. **Progressive Disclosure**: Metadata always loaded, instructions on activation, resources on demand
2. **Industry Detection**: Auto-detect SaaS, e-commerce, local, publisher, agency
3. **Parallel Execution**: Full audits spawn 7 subagents simultaneously
-4. **Extension System**: DataForSEO MCP integration for live data
+4. **Extension System**: DataForSEO MCP for live data, Banana MCP for AI image generation
diff --git a/docs/ARCHITECTURE.md b/docs/ARCHITECTURE.md
index d5bb003..538468c 100644
--- a/docs/ARCHITECTURE.md
+++ b/docs/ARCHITECTURE.md
@@ -198,20 +198,48 @@ Extensions are opt-in add-ons that integrate external data sources via MCP serve
```
extensions/
-└── dataforseo/ # DataForSEO MCP integration
+├── dataforseo/ # DataForSEO MCP integration
+│ ├── README.md # Extension documentation
+│ ├── install.sh # Unix installer
+│ ├── install.ps1 # Windows installer
+│ ├── uninstall.sh # Unix uninstaller
+│ ├── uninstall.ps1 # Windows uninstaller
+│ ├── field-config.json # API response field filtering
+│ ├── skills/
+│ │ └── seo-dataforseo/
+│ │ └── SKILL.md # Sub-skill (22 commands)
+│ ├── agents/
+│ │ └── seo-dataforseo.md # Subagent
+│ └── docs/
+│ └── DATAFORSEO-SETUP.md # Account setup guide
+│
+└── banana/ # Banana Image Generation (Gemini AI)
├── README.md # Extension documentation
├── install.sh # Unix installer
- ├── install.ps1 # Windows installer
├── uninstall.sh # Unix uninstaller
- ├── uninstall.ps1 # Windows uninstaller
- ├── field-config.json # API response field filtering
├── skills/
- │ └── seo-dataforseo/
- │ └── SKILL.md # Sub-skill (22 commands)
+ │ └── seo-image-gen/
+ │ └── SKILL.md # Sub-skill (6 commands)
├── agents/
- │ └── seo-dataforseo.md # Subagent
+ │ └── seo-image-gen.md # Image audit subagent
+ ├── scripts/ # Python scripts (stdlib only)
+ │ ├── generate.py # Direct API fallback
+ │ ├── edit.py # Image editing fallback
+ │ ├── batch.py # CSV batch workflow
+ │ ├── cost_tracker.py # Usage and cost tracking
+ │ ├── presets.py # Brand preset management
+ │ ├── setup_mcp.py # MCP configuration
+ │ └── validate_setup.py # Installation verification
+ ├── references/ # On-demand knowledge
+ │ ├── prompt-engineering.md # 6-component Reasoning Brief
+ │ ├── gemini-models.md # Model specs and pricing
+ │ ├── mcp-tools.md # MCP tool reference
+ │ ├── post-processing.md # ImageMagick recipes
+ │ ├── cost-tracking.md # Cost tracking guide
+ │ ├── presets.md # Preset schema
+ │ └── seo-image-presets.md # SEO-specific presets
└── docs/
- └── DATAFORSEO-SETUP.md # Account setup guide
+ └── BANANA-SETUP.md # API key and MCP setup
```
### Available Extensions
@@ -219,6 +247,7 @@ extensions/
| Extension | Package | What it Adds |
|-----------|---------|-------------|
| **DataForSEO** | `dataforseo-mcp-server` | 22 commands: live SERP, keywords, backlinks, on-page analysis, content analysis, business listings, AI visibility, LLM mentions |
+| **Banana Image Gen** | `@ycse/nanobanana-mcp` | 6 commands: OG image, hero image, product photo, infographic, custom, and batch generation via Gemini AI |
### Extension Convention
diff --git a/docs/COMMANDS.md b/docs/COMMANDS.md
index a7f4ba8..8f5c1a0 100644
--- a/docs/COMMANDS.md
+++ b/docs/COMMANDS.md
@@ -314,6 +314,37 @@ Live SEO data via DataForSEO MCP server (extension). 22 commands across 9 API mo
---
+### `/seo image-gen [use-case] <description>`
+
+AI image generation for SEO assets (extension). Powered by Gemini via nanobanana-mcp.
+
+**Prerequisites:** Banana extension installed (`./extensions/banana/install.sh`)
+
+**Use Cases:**
+```
+/seo image-gen og <description> # OG/social preview image (16:9, 1K)
+/seo image-gen hero <description> # Blog hero image (16:9, 2K)
+/seo image-gen product <description> # Product photography (4:3, 2K)
+/seo image-gen infographic <description> # Infographic visual (2:3, 4K)
+/seo image-gen custom <description> # Custom with full Creative Director pipeline
+/seo image-gen batch <description> [N] # Generate N variations (default: 3)
+```
+
+**Example:**
+```
+/seo image-gen og "Professional SaaS analytics dashboard with clean UI"
+/seo image-gen hero "Dramatic sunset over modern city skyline"
+/seo image-gen product "Wireless noise-canceling headphones on marble surface"
+```
+
+**What it does:**
+1. Maps SEO use case to optimized domain mode, aspect ratio, and resolution
+2. Constructs 6-component Reasoning Brief (Creative Director pipeline)
+3. Generates image via Gemini API
+4. Provides SEO checklist (alt text, file naming, WebP, schema markup)
+
+---
+
## Quick Reference
| Command | Use Case |
@@ -324,6 +355,7 @@ Live SEO data via DataForSEO MCP server (extension). 22 commands across 9 API mo
| `/seo geo <url>` | AI search optimization |
| `/seo hreflang [url]` | Hreflang/i18n SEO audit |
| `/seo images <url>` | Image optimization |
+| `/seo image-gen [use-case] <desc>` | AI image generation (extension) |
| `/seo page <url>` | Single page analysis |
| `/seo plan <type>` | Strategic planning |
| `/seo programmatic [url\|plan]` | Programmatic SEO analysis |
diff --git a/extensions/banana/README.md b/extensions/banana/README.md
new file mode 100644
index 0000000..c247410
--- /dev/null
+++ b/extensions/banana/README.md
@@ -0,0 +1,92 @@
+# Banana Image Generation Extension for Claude SEO
+
+Generate production-ready SEO images using AI — OG/social previews, blog heroes,
+product photography, infographics, and more. Powered by Google Gemini via the
+banana Creative Director pipeline.
+
+## Prerequisites
+
+- **Claude SEO** installed (`~/.claude/skills/seo/`)
+- **Node.js 18+** with npx
+- **Google AI API key** (free at [aistudio.google.com/apikey](https://aistudio.google.com/apikey))
+- **ImageMagick** (optional, for post-processing)
+
+## Installation
+
+```bash
+./extensions/banana/install.sh
+```
+
+The installer will:
+1. Verify Claude SEO is installed
+2. Prompt for your Google AI API key (if nanobanana-mcp not already configured)
+3. Install the `seo-image-gen` skill and agent
+4. Configure the MCP server in `~/.claude/settings.json`
+
+## Commands
+
+| Command | What it does |
+|---------|-------------|
+| `/seo image-gen og <description>` | OG/social preview image (1200x630 feel) |
+| `/seo image-gen hero <description>` | Blog hero image (widescreen, dramatic) |
+| `/seo image-gen product <description>` | Product photography (clean, white BG) |
+| `/seo image-gen infographic <description>` | Infographic visual (vertical, data-heavy) |
+| `/seo image-gen custom <description>` | Custom with full Creative Director pipeline |
+| `/seo image-gen batch <description> [N]` | Generate N variations (default: 3) |
+
+## Use Case Defaults
+
+| Use Case | Aspect Ratio | Resolution | Domain Mode | Cost |
+|----------|-------------|------------|-------------|------|
+| OG/Social Preview | 16:9 | 1K | Product/UI | ~$0.04 |
+| Blog Hero | 16:9 | 2K | Cinema/Editorial | ~$0.08 |
+| Product Photo | 4:3 | 2K | Product | ~$0.08 |
+| Infographic | 2:3 | 4K | Infographic | ~$0.16 |
+| Social Square | 1:1 | 1K | UI/Web | ~$0.04 |
+| Favicon/Icon | 1:1 | 512 | Logo | ~$0.02 |
+
+## How It Works
+
+Claude acts as a **Creative Director** — it never passes raw text to the API.
+Instead, it analyzes your intent, selects the optimal domain mode, and constructs
+an optimized prompt using a proven 6-component Reasoning Brief system:
+
+1. **Subject** (30%) — Physical specificity and micro-details
+2. **Style** (25%) — Camera specs, film stock, brand references
+3. **Context** (15%) — Location, time, weather, supporting elements
+4. **Action** (10%) — Pose, gesture, movement, state
+5. **Composition** (10%) — Shot type, framing, focal length
+6. **Lighting** (10%) — Direction, quality, color temperature
+
+## Post-Generation SEO Checklist
+
+After every generation, Claude provides:
+- Alt text suggestion (keyword-rich, descriptive)
+- SEO-friendly file naming convention
+- WebP conversion command
+- ImageObject schema snippet
+- OG meta tag markup (for social previews)
+
+## Audit Integration
+
+During `/seo audit`, the extension optionally spawns an image analysis agent that:
+- Audits existing OG/social images across the site
+- Identifies missing or low-quality images
+- Creates a prioritized generation plan with prompt suggestions
+- Estimates total cost for the generation plan
+
+The agent never auto-generates images — it produces a plan for your review.
+
+## Uninstallation
+
+```bash
+./extensions/banana/uninstall.sh
+```
+
+This removes the skill and agent. If you also use the standalone banana skill,
+the MCP server config is preserved.
+
+## Troubleshooting
+
+See [docs/BANANA-SETUP.md](docs/BANANA-SETUP.md) for detailed setup instructions
+and common issues.
diff --git a/extensions/banana/agents/seo-image-gen.md b/extensions/banana/agents/seo-image-gen.md
new file mode 100644
index 0000000..0b09bc9
--- /dev/null
+++ b/extensions/banana/agents/seo-image-gen.md
@@ -0,0 +1,59 @@
+---
+name: seo-image-gen
+description: SEO image analyst. Audits existing OG/social preview images, identifies missing or low-quality images, and creates an image generation plan with prompts for key pages. Does NOT auto-generate images.
+tools: Read, Bash, Glob, Grep
+---
+
+You are an SEO image analyst. When delegated tasks during an SEO audit:
+
+1. Check that nanobanana-mcp tools are available before including generation recommendations
+2. Analyze the site's existing image strategy for SEO impact
+3. Output a structured generation plan — never auto-generate (cost control)
+
+## Analysis Scope
+
+For each audited page, evaluate:
+- **OG image presence** — Does `og:image` meta tag exist? Is it valid?
+- **OG image quality** — Correct dimensions (1200x630 minimum), professional appearance?
+- **Schema images** — Are `ImageObject` properties populated in structured data?
+- **Alt text quality** — Descriptive, keyword-rich, not stuffed?
+- **Image format** — Using modern formats (WebP, AVIF) vs legacy (PNG, JPEG)?
+- **Image file size** — Under 200KB for hero, under 100KB for thumbnails?
+
+## Output Format
+
+Match existing claude-seo patterns:
+
+### Image Audit Summary
+
+| Metric | Value | Status |
+|--------|-------|--------|
+| Pages with OG images | X/Y | Pass/Fail |
+| OG images correct size | X/Y | Pass/Fail |
+| Schema ImageObject usage | X/Y | Pass/Fail |
+| WebP/AVIF adoption | X% | Pass/Fail |
+| Average image file size | XKB | Pass/Fail |
+
+### Image Generation Plan
+
+For each page missing or having low-quality images:
+
+| Page | Issue | Suggested Use Case | Prompt Idea | Priority |
+|------|-------|-------------------|-------------|----------|
+| /homepage | Missing OG image | og | Professional SaaS dashboard overview | Critical |
+| /blog/post-1 | Low-res hero | hero | [contextual suggestion] | High |
+
+Priority levels: Critical > High > Medium > Low
+
+### Recommendations
+
+- Prioritize pages by traffic volume (highest traffic = fix first)
+- Note estimated cost for full generation plan
+- Suggest batch generation for efficiency
+- Recommend WebP conversion pipeline for all generated assets
+
+## Error Handling
+
+- If nanobanana-mcp is not available, still audit existing images but note that generation requires the banana extension
+- Report errors clearly with actionable next steps
+- Note data source as "Image Audit (static analysis)" to distinguish from live checks
diff --git a/extensions/banana/docs/BANANA-SETUP.md b/extensions/banana/docs/BANANA-SETUP.md
new file mode 100644
index 0000000..9502ab7
--- /dev/null
+++ b/extensions/banana/docs/BANANA-SETUP.md
@@ -0,0 +1,86 @@
+# Banana Extension Setup Guide
+
+## Google AI API Key
+
+1. Go to [aistudio.google.com/apikey](https://aistudio.google.com/apikey)
+2. Sign in with your Google account
+3. Click "Create API key"
+4. Copy the key — you'll need it during installation
+
+**Free tier limits:**
+- ~10 requests per minute (RPM)
+- ~500 requests per day (RPD)
+- Resets at midnight Pacific time
+
+## MCP Server Configuration
+
+The installer configures this automatically. If you need to set it up manually,
+add to `~/.claude/settings.json`:
+
+```json
+{
+ "mcpServers": {
+ "nanobanana-mcp": {
+ "command": "npx",
+ "args": ["-y", "@ycse/nanobanana-mcp@latest"],
+ "env": {
+ "GOOGLE_API_KEY": "your-api-key-here"
+ }
+ }
+ }
+}
+```
+
+## Verifying Installation
+
+Run the validation script:
+```bash
+python3 ~/.claude/skills/seo-image-gen/scripts/validate_setup.py
+```
+
+Or check manually:
+1. `ls ~/.claude/skills/seo-image-gen/SKILL.md` — skill file exists
+2. `ls ~/.claude/agents/seo-image-gen.md` — agent file exists
+3. `grep nanobanana ~/.claude/settings.json` — MCP configured
+
+## Common Issues
+
+### "MCP tools not available"
+- Restart Claude Code after installing the extension
+- Verify your API key is valid at [aistudio.google.com](https://aistudio.google.com)
+- Check `~/.claude/settings.json` has the nanobanana-mcp entry
+
+### "Rate limited (429)"
+- Free tier: ~10 requests/minute, ~500/day
+- Wait 60 seconds and retry
+- For batch operations, add delays between requests
+
+### "IMAGE_SAFETY" error
+- The safety filter flagged your prompt (often a false positive)
+- Claude will suggest rephrased alternatives automatically
+- Common triggers: certain color descriptions, implied scenarios
+- See `references/prompt-engineering.md` Safety Rephrase section
+
+### "Node.js version too old"
+- Requires Node.js 18+
+- Update via nvm: `nvm install 18 && nvm use 18`
+- Or download from [nodejs.org](https://nodejs.org/)
+
+### Generated images not appearing
+- Default output directory: `~/Documents/nanobanana_generated/`
+- Check the path returned by Claude after generation
+- Verify disk space is available
+
+## ImageMagick (Optional)
+
+For post-processing (WebP conversion, cropping, background removal):
+
+```bash
+# Ubuntu/Pop!_OS
+sudo apt install imagemagick
+
+# Verify
+magick --version
+```
+
+If `magick` (v7) is not available, the scripts fall back to `convert` (v6).
diff --git a/extensions/banana/install.sh b/extensions/banana/install.sh
new file mode 100755
index 0000000..d165cf5
--- /dev/null
+++ b/extensions/banana/install.sh
@@ -0,0 +1,170 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+# Banana Image Generation Extension Installer for Claude SEO
+# Wraps everything in main() to prevent partial execution on network failure
+
+main() {
+ SKILL_DIR="${HOME}/.claude/skills/seo-image-gen"
+ AGENT_DIR="${HOME}/.claude/agents"
+ SEO_SKILL_DIR="${HOME}/.claude/skills/seo"
+ SETTINGS_FILE="${HOME}/.claude/settings.json"
+
+ echo "════════════════════════════════════════"
+ echo "║ Banana Image Gen - SEO Extension ║"
+ echo "║ For Claude SEO ║"
+ echo "════════════════════════════════════════"
+ echo ""
+
+ # Check prerequisites
+ if [ ! -d "${SEO_SKILL_DIR}" ]; then
+ echo "✗ Claude SEO is not installed."
+ echo " Install it first: curl -fsSL https://raw.githubusercontent.com/AgriciDaniel/claude-seo/main/install.sh | bash"
+ exit 1
+ fi
+ echo "✓ Claude SEO detected"
+
+ if ! command -v node >/dev/null 2>&1; then
+ echo "✗ Node.js is required but not installed."
+ echo " Install Node.js 20+: https://nodejs.org/"
+ exit 1
+ fi
+
+ NODE_VERSION=$(node -v | sed 's/v//' | cut -d. -f1)
+ if [ "${NODE_VERSION}" -lt 20 ]; then
+ echo "✗ Node.js 20+ required (found v${NODE_VERSION})."
+ echo " Update: https://nodejs.org/"
+ exit 1
+ fi
+ echo "✓ Node.js v$(node -v | sed 's/v//') detected"
+
+ if ! command -v npx >/dev/null 2>&1; then
+ echo "✗ npx is required but not found (comes with npm)."
+ exit 1
+ fi
+ echo "✓ npx detected"
+
+ # Determine script directory (works for both ./install.sh and repo-relative paths)
+ SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+
+ # Check if running from repo or standalone
+ if [ -f "${SCRIPT_DIR}/skills/seo-image-gen/SKILL.md" ]; then
+ SOURCE_DIR="${SCRIPT_DIR}"
+ elif [ -f "${SCRIPT_DIR}/extensions/banana/skills/seo-image-gen/SKILL.md" ]; then
+ SOURCE_DIR="${SCRIPT_DIR}/extensions/banana"
+ else
+ echo "✗ Cannot find extension source files."
+ echo " Run this script from the claude-seo repo: ./extensions/banana/install.sh"
+ exit 1
+ fi
+
+ # Check if nanobanana-mcp is already configured
+ MCP_CONFIGURED=false
+ if [ -f "${SETTINGS_FILE}" ]; then
+ if python3 -c "
+import json
+with open('${SETTINGS_FILE}', 'r') as f:
+ settings = json.load(f)
+if 'mcpServers' in settings and 'nanobanana-mcp' in settings['mcpServers']:
+ exit(0)
+else:
+ exit(1)
+" 2>/dev/null; then
+ MCP_CONFIGURED=true
+ echo "✓ nanobanana-mcp already configured in settings.json"
+ fi
+ fi
+
+ # If MCP not configured, prompt for API key
+ if [ "${MCP_CONFIGURED}" = false ]; then
+ echo ""
+ echo "Google AI API key required for image generation."
+ echo "Get a free key at: https://aistudio.google.com/apikey"
+ echo ""
+
+ read -rsp "Google AI API key (GOOGLE_AI_API_KEY): " GOOGLE_AI_API_KEY
+ echo ""
+ if [ -z "${GOOGLE_AI_API_KEY}" ]; then
+ echo "✗ API key cannot be empty."
+ exit 1
+ fi
+
+ # Configure MCP server
+ echo "→ Configuring nanobanana-mcp server..."
+ python3 -c "
+import json, os
+
+settings_path = '${SETTINGS_FILE}'
+
+# Read existing settings or create new
+if os.path.exists(settings_path):
+ with open(settings_path, 'r') as f:
+ settings = json.load(f)
+else:
+ settings = {}
+
+# Ensure mcpServers key exists
+if 'mcpServers' not in settings:
+ settings['mcpServers'] = {}
+
+# Add nanobanana-mcp server config
+settings['mcpServers']['nanobanana-mcp'] = {
+ 'command': 'npx',
+ 'args': ['-y', '@ycse/nanobanana-mcp@latest'],
+ 'env': {
+ 'GOOGLE_AI_API_KEY': '''${GOOGLE_AI_API_KEY}'''
+ }
+}
+
+# Write back
+os.makedirs(os.path.dirname(settings_path), exist_ok=True)
+with open(settings_path, 'w') as f:
+ json.dump(settings, f, indent=2)
+
+print(' ✓ nanobanana-mcp configured in settings.json')
+" || {
+ echo "✗ Could not auto-configure MCP server."
+ echo " See: extensions/banana/docs/BANANA-SETUP.md"
+ exit 1
+ }
+ fi
+
+ # Install skill
+ echo ""
+ echo "→ Installing seo-image-gen skill..."
+ mkdir -p "${SKILL_DIR}"
+ cp "${SOURCE_DIR}/skills/seo-image-gen/SKILL.md" "${SKILL_DIR}/SKILL.md"
+
+ # Install agent
+ echo "→ Installing seo-image-gen agent..."
+ mkdir -p "${AGENT_DIR}"
+ cp "${SOURCE_DIR}/agents/seo-image-gen.md" "${AGENT_DIR}/seo-image-gen.md"
+
+ # Copy scripts and references to skill directory for ${CLAUDE_SKILL_DIR} resolution
+ echo "→ Installing scripts and references..."
+ mkdir -p "${SKILL_DIR}/scripts" "${SKILL_DIR}/references"
+ cp "${SOURCE_DIR}/scripts/"*.py "${SKILL_DIR}/scripts/"
+ cp "${SOURCE_DIR}/references/"*.md "${SKILL_DIR}/references/"
+
+ # Pre-warm npx package
+ echo "→ Pre-downloading nanobanana-mcp..."
+ npx -y @ycse/nanobanana-mcp@latest --help >/dev/null 2>&1 || true
+
+ echo ""
+ echo "✓ Banana Image Generation extension installed successfully!"
+ echo ""
+ echo "Usage:"
+ echo " 1. Start Claude Code: claude"
+ echo " 2. Run commands:"
+ echo " /seo image-gen og \"Professional SaaS dashboard\""
+ echo " /seo image-gen hero \"Dramatic sunset over city skyline\""
+ echo " /seo image-gen product \"Wireless headphones on marble\""
+ echo " /seo image-gen infographic \"SEO ranking factors 2026\""
+ echo " /seo image-gen custom \"Any creative concept\""
+ echo " /seo image-gen batch \"Product variations\" 3"
+ echo ""
+ echo "Full docs: extensions/banana/README.md"
+ echo "To uninstall: ./extensions/banana/uninstall.sh"
+}
+
+main "$@"
diff --git a/extensions/banana/references/cost-tracking.md b/extensions/banana/references/cost-tracking.md
new file mode 100644
index 0000000..80e89d4
--- /dev/null
+++ b/extensions/banana/references/cost-tracking.md
@@ -0,0 +1,47 @@
+# Cost Tracking Reference
+
+> Load this on-demand when the user asks about costs or before batch operations.
+
+## Pricing Table
+
+| Model | Resolution | Cost/Image | Notes |
+|-------|-----------|-----------|-------|
+| 3.1 Flash | 512 | $0.020 | Quick drafts |
+| 3.1 Flash | 1K | $0.039 | Standard (default) |
+| 3.1 Flash | 2K | $0.078 | Quality assets |
+| 3.1 Flash | 4K | $0.156 | Print/hero images |
+| 2.5 Flash | 512 | $0.020 | Draft fallback |
+| 2.5 Flash | 1K | $0.039 | Standard fallback |
+| Batch API | Any | 50% of above | Asynchronous, higher latency |
+
+Pricing is approximate, based on ~1,290 output tokens per image.
+Research suggests actual costs may be ~$0.067/img. Verify at https://ai.google.dev/gemini-api/docs/pricing
+
+## Free Tier Limits
+
+- ~10 requests per minute (RPM)
+- ~500 requests per day (RPD)
+- Per Google Cloud project, resets midnight Pacific
+
+## Cost Tracker Commands
+
+```bash
+# Log a generation
+cost_tracker.py log --model gemini-3.1-flash-image-preview --resolution 1K --prompt "coffee shop hero"
+
+# View summary (total + last 7 days)
+cost_tracker.py summary
+
+# Today's usage
+cost_tracker.py today
+
+# Estimate before batch
+cost_tracker.py estimate --model gemini-3.1-flash-image-preview --resolution 1K --count 10
+
+# Reset ledger
+cost_tracker.py reset --confirm
+```
+
+## Storage
+
+Ledger stored at `~/.banana/costs.json`. Created automatically on first use.
diff --git a/extensions/banana/references/gemini-models.md b/extensions/banana/references/gemini-models.md
new file mode 100644
index 0000000..f8e2640
--- /dev/null
+++ b/extensions/banana/references/gemini-models.md
@@ -0,0 +1,200 @@
+# Gemini Image Generation Models
+
+> Last updated: 2026-03-13
+> Aligned with Google's March 2026 API state
+
+## Available Models
+
+### gemini-3.1-flash-image-preview (Recommended)
+| Property | Value |
+|----------|-------|
+| **Model ID** | `gemini-3.1-flash-image-preview` |
+| **Tier** | Nano Banana 2 (Flash) |
+| **Speed** | Fast — optimized for high-volume use |
+| **Aspect Ratios** | All 14 ratios (see table below) |
+| **Max Resolution** | Up to 4096×4096 (4K tier) |
+| **Features** | Google Search grounding (web + image), thinking levels, image-only output, extreme aspect ratios |
+| **Rate Limits (Free)** | ~10 RPM / ~500 RPD (per Google Cloud project, resets midnight Pacific) |
+| **Output Tokens** | ~1,290 output tokens per image |
+| **Best For** | Most use cases, rapid iteration, batch generation |
+
+### gemini-2.5-flash-image
+| Property | Value |
+|----------|-------|
+| **Model ID** | `gemini-2.5-flash-image` |
+| **Tier** | Nano Banana 2 (Flash, previous gen) |
+| **Speed** | Fast |
+| **Aspect Ratios** | 1:1, 16:9, 9:16, 4:3, 3:4 |
+| **Max Resolution** | Up to 1024×1024 (1K tier) |
+| **Rate Limits (Free)** | ~10 RPM / ~500 RPD |
+| **Best For** | Stable fallback, proven quality |
+
+## Deprecated Models (DO NOT USE)
+
+### gemini-3-pro-image-preview
+- **Status:** Base model deprecated March 9, 2026. **Image generation variant may still be accessible** — use at your own discretion via `set_model`. Prefer 3.1 Flash.
+- **Was:** Nano Banana Pro tier (professional asset production, 4K output, 14 reference images)
+- **Migration:** Use `gemini-3.1-flash-image-preview` instead
+
+### gemini-2.0-flash-exp
+- **Status:** Deprecated, replaced by gemini-2.5-flash-image
+
+## Aspect Ratios
+
+All 14 supported ratios. Availability varies by model:
+
+| Ratio | Orientation | Use Cases | 3.1 Flash | 2.5 Flash |
+|-------|-------------|-----------|:---------:|:---------:|
+| `1:1` | Square | Social posts, avatars, thumbnails | ✅ | ✅ |
+| `16:9` | Landscape | Blog headers, YouTube thumbnails, presentations | ✅ | ✅ |
+| `9:16` | Portrait | Stories, Reels, TikTok, mobile | ✅ | ✅ |
+| `4:3` | Landscape | Product shots, classic display | ✅ | ✅ |
+| `3:4` | Portrait | Book covers, portrait framing | ✅ | ✅ |
+| `2:3` | Portrait | Pinterest pins, posters | ✅ | ❌ |
+| `3:2` | Landscape | DSLR standard, photo prints | ✅ | ❌ |
+| `4:5` | Portrait | Instagram portrait, social | ✅ | ❌ |
+| `5:4` | Landscape | Large format photography | ✅ | ❌ |
+| `1:4` | Tall strip | Vertical banners, side panels | ✅ | ❌ |
+| `4:1` | Wide strip | Website banners, headers | ✅ | ❌ |
+| `1:8` | Extreme tall | Narrow vertical strips | ✅ | ❌ |
+| `8:1` | Extreme wide | Ultra-wide banners | ✅ | ❌ |
+| `21:9` | Ultra-wide | Cinematic, film-grade, ultra-wide monitors | ✅ | ❌ |
+
+## Resolution Tiers
+
+Control output resolution with the `imageSize` parameter. Note the **uppercase K** requirement.
+
+| `imageSize` Value | Pixel Range | Model Availability | Use Case |
+|-------------------|-------------|-------------------|----------|
+| `512` | Up to 512×512 | All models | Drafts, quick iteration, low bandwidth |
+| `1K` (default) | Up to 1024×1024 | All models | Standard web use, social media |
+| `2K` | Up to 2048×2048 | 3.1 Flash | Quality assets, detailed work |
+| `4K` | Up to 4096×4096 | 3.1 Flash | Print production, hero images, final assets |
+
+**Notes:**
+- Actual pixel dimensions depend on aspect ratio (e.g., 4K at 16:9 = 4096×2304)
+- Higher resolutions consume more tokens and cost more
+- Default is `1K` if `imageSize` is not specified
+
+## API Configuration
+
+### Endpoint
+```
+https://generativelanguage.googleapis.com/v1beta/models/{model-id}:generateContent
+```
+
+### Required Parameters
+```json
+{
+ "contents": [{"parts": [{"text": "your prompt here"}]}],
+ "generationConfig": {
+ "responseModalities": ["TEXT", "IMAGE"],
+ "imageConfig": {
+ "aspectRatio": "16:9",
+ "imageSize": "1K"
+ }
+ }
+}
+```
+
+### Image-Only Output Mode
+Force the model to return only an image (no text response):
+```json
+{
+ "generationConfig": {
+ "responseModalities": ["IMAGE"]
+ }
+}
+```
+
+### Thinking Level
+Control how much the model "thinks" before generating. Higher levels improve complex compositions but increase latency:
+```json
+{
+ "generationConfig": {
+ "thinkingConfig": {
+ "thinkingLevel": "medium"
+ }
+ }
+}
+```
+Levels: `minimal`, `low`, `medium`, `high`
+
+### Google Search Grounding
+Ground generation in real-world visual references. Supports web and image search (3.1 Flash):
+```json
+{
+ "tools": [{"googleSearch": {}}]
+}
+```
+**Prompt pattern:** `[Search/source request] + [Analytical task] + [Visual translation]`
+
+Example: "Search for the latest SpaceX Starship design, analyze its proportions and markings, then generate a photorealistic image of it at sunset on the launch pad."
+
+### Multi-Image Input
+Up to 14 reference images can be provided:
+- **10 object references** — for style, composition, or visual matching
+- **4 character references** — assign distinct names to preserve features across generations
+
+Useful for character consistency, style transfer, and brand-aligned generation.
+
+## Rate Limits by Tier
+
+| Tier | RPM | RPD | Notes |
+|------|-----|-----|-------|
+| Free | ~10 | ~500 | Per Google Cloud project, resets midnight Pacific. Reduced Dec 2025. |
+| Pay-as-you-go | 30 | 10,000 | Production workloads |
+| Enterprise | Custom | Custom | Contact Google |
+
+## Pricing
+
+| Model | Resolution | Cost per Image | Notes |
+|-------|-----------|---------------|-------|
+| 3.1 Flash | 1K | ~$0.039 | Standard |
+| 3.1 Flash | 2K | ~$0.078 | 2× standard |
+| 3.1 Flash | 4K | ~$0.156 | 4× standard |
+| 2.5 Flash | 1K | ~$0.039 | Standard |
+| Batch API | Any | 50% discount | Asynchronous, higher latency |
+
+Pricing is approximate and based on ~1,290 output tokens per image.
+Research suggests NB2 pricing may be ~$0.067/img (vs documented $0.039). Verify current pricing at https://ai.google.dev/gemini-api/docs/pricing
+
+## Image Output Specs
+
+| Property | Value |
+|----------|-------|
+| **Format** | PNG |
+| **Max Resolution** | Up to 4096×4096 (4K tier, 3.1 Flash) |
+| **Color Space** | sRGB |
+| **Text Rendering** | Supported — best under 25 characters |
+| **Style Control** | Via prompt engineering |
+
+## Safety Filters
+
+Gemini uses a two-layer safety architecture:
+
+1. **Input filters** — block prompts containing prohibited content before generation
+2. **Output filters** — analyze generated images and block unsafe results
+
+| `finishReason` | Meaning | Retryable? |
+|----------------|---------|:----------:|
+| `STOP` | Successful generation | N/A |
+| `IMAGE_SAFETY` | Output blocked by safety filter | Rephrase prompt |
+| `PROHIBITED_CONTENT` | Content policy violation | No — topic is blocked |
+| `SAFETY` | General safety block | Rephrase prompt |
+| `RECITATION` | Detected copyrighted content | Rephrase prompt |
+
+**Known issue:** Filters are known to be overly cautious — benign prompts may be blocked. Iterate with rephrased wording if this happens.
+
+## Content Credentials
+
+- **SynthID watermarks** are always embedded in generated images (invisible, machine-readable)
+- **C2PA metadata** is included on Pro/paid outputs (verifiable provenance chain)
+
+## Key Limitations
+- No video generation (image only)
+- No transparent backgrounds (PNG but always with background)
+- Text rendering quality varies — keep text under 25 characters for best results
+- Safety filters may block some prompts (violence, NSFW, public figures) — known to be overly cautious
+- Session context resets between Claude Code conversations
+- `imageSize` and thinking level depend on MCP package version support
diff --git a/extensions/banana/references/mcp-tools.md b/extensions/banana/references/mcp-tools.md
new file mode 100644
index 0000000..c1c3b56
--- /dev/null
+++ b/extensions/banana/references/mcp-tools.md
@@ -0,0 +1,115 @@
+# MCP Tools Reference — @ycse/nanobanana-mcp
+
+> Package: `@ycse/nanobanana-mcp`
+> GitHub: https://github.com/YCSE/nanobanana-mcp
+
+## Tools
+
+### gemini_generate_image
+Generate an image from a text prompt.
+
+**Parameters:**
+| Param | Type | Required | Description |
+|-------|------|----------|-------------|
+| `prompt` | string | Yes | Text description of the image to generate |
+
+**Returns:** Image data + file path (saved to `~/Documents/nanobanana_generated/`)
+
+**Example usage in Claude Code:**
+```
+User: "Generate a sunset over mountains in watercolor style"
+→ Claude calls gemini_generate_image with prompt
+→ Returns image path and description
+```
+
+### gemini_edit_image
+Edit an existing image with text instructions.
+
+**Parameters:**
+| Param | Type | Required | Description |
+|-------|------|----------|-------------|
+| `imagePath` | string | Yes | Path to the image file to edit |
+| `prompt` | string | Yes | Edit instructions |
+
+**Returns:** Modified image data + file path
+
+**Example:**
+```
+User: "Remove the background from ~/Documents/photo.png"
+→ Claude calls gemini_edit_image with path and instruction
+```
+
+### gemini_chat
+Multi-turn visual conversation maintaining session context.
+
+**Parameters:**
+| Param | Type | Required | Description |
+|-------|------|----------|-------------|
+| `message` | string | Yes | Chat message (can reference previous images) |
+
+**Returns:** Text response + optional image
+
+**Key feature:** Session consistency — maintains style, characters, and context across turns. Great for iterative refinement.
+
+### set_aspect_ratio
+Configure the aspect ratio for subsequent image generations.
+
+**Parameters:**
+| Param | Type | Required | Description |
+|-------|------|----------|-------------|
+| `ratio` | string | Yes | Aspect ratio (e.g., "16:9", "1:1", "9:16") |
+
+**Supported ratios:** 1:1, 16:9, 9:16, 4:3, 3:4, 2:3, 3:2, 4:5, 5:4, 1:4, 4:1, 1:8, 8:1, 21:9
+
+### set_model
+Switch the active Gemini model.
+
+**Parameters:**
+| Param | Type | Required | Description |
+|-------|------|----------|-------------|
+| `model` | string | Yes | Model identifier |
+
+**Available models:**
+- `gemini-3.1-flash-image-preview` (default, recommended)
+- `gemini-2.5-flash-image` (stable fallback)
+
+### get_image_history
+Retrieve list of images generated in the current session.
+
+**Parameters:** None
+
+**Returns:** Array of image entries with paths and prompts
+
+### clear_conversation
+Reset session context and conversation history.
+
+**Parameters:** None
+
+**Returns:** Confirmation of reset
+
+## Environment Variables
+
+| Variable | Required | Description |
+|----------|----------|-------------|
+| `GOOGLE_AI_API_KEY` | Yes | API key from https://aistudio.google.com/apikey |
+| `NANOBANANA_MODEL` | No | Override default model (default: `gemini-3.1-flash-image-preview`) |
+
+## Output Directory
+All generated images are saved to: `~/Documents/nanobanana_generated/`
+
+Images are named with timestamps for easy identification.
+
+## Feature Availability via MCP
+
+Some newer Gemini API features depend on the MCP package version of `@ycse/nanobanana-mcp`. Check the package version to confirm support:
+
+| Feature | API Status | MCP Support |
+|---------|-----------|-------------|
+| `imageSize` (resolution control) | Available | Depends on package version |
+| Thinking level (`thinkingConfig`) | Available | Depends on package version |
+| Search grounding (`googleSearch`) | Available | Depends on package version |
+| Image-only output (`responseModalities: ["IMAGE"]`) | Available | Depends on package version |
+| Multi-image input (up to 14 refs) | Available | Via `gemini_chat` with image paths |
+| All 14 aspect ratios | Available | Via `set_aspect_ratio` |
+
+If a feature is not yet supported by the MCP package, you can still use it via direct API calls with `curl` or the Google AI SDK.
diff --git a/extensions/banana/references/post-processing.md b/extensions/banana/references/post-processing.md
new file mode 100644
index 0000000..546ad5c
--- /dev/null
+++ b/extensions/banana/references/post-processing.md
@@ -0,0 +1,192 @@
+# Post-Processing Pipeline Reference
+
+> Load this on-demand when the user needs image manipulation after generation.
+
+## Prerequisites
+
+Check availability before using:
+```bash
+which magick # ImageMagick 7 (preferred)
+which convert # ImageMagick 6 (fallback)
+which ffmpeg # For video/animation
+```
+
+Install ImageMagick if not present: `sudo apt install imagemagick` (Debian/Ubuntu) or `brew install imagemagick` (macOS).
+
+## Common Operations
+
+### Resize for Platforms
+
+```bash
+# Instagram post (1080x1080)
+magick input.png -resize 1080x1080^ -gravity center -extent 1080x1080 instagram.png
+
+# Twitter/X header (1500x500)
+magick input.png -resize 1500x500^ -gravity center -extent 1500x500 twitter-header.png
+
+# YouTube thumbnail (1280x720)
+magick input.png -resize 1280x720^ -gravity center -extent 1280x720 youtube-thumb.png
+
+# LinkedIn banner (1584x396)
+magick input.png -resize 1584x396^ -gravity center -extent 1584x396 linkedin-banner.png
+
+# Favicon (multi-size ICO)
+magick input.png -resize 32x32 favicon.ico
+```
+
+### Background Removal (Transparency)
+
+```bash
+# Remove solid white background
+magick input.png -fuzz 10% -transparent white output.png
+
+# Remove solid color background (specify color)
+magick input.png -fuzz 15% -transparent "#F0F0F0" output.png
+
+# Clean edges after transparency (anti-alias)
+magick input.png -fuzz 10% -transparent white -channel A -blur 0x1 -level 50%,100% output.png
+
+# Auto-crop transparent padding
+magick input.png -trim +repage output.png
+```
+
+### Format Conversion
+
+```bash
+# PNG to WebP (web-optimized, smaller file)
+magick input.png -quality 85 output.webp
+
+# PNG to JPEG (with white background for transparency)
+magick input.png -background white -flatten -quality 90 output.jpg
+
+# PNG to AVIF (modern, smallest size)
+magick input.png -quality 80 output.avif
+
+# SVG trace (for logos — requires potrace)
+potrace input.pbm -s -o output.svg
+```
+
+### Color Adjustments
+
+```bash
+# Increase contrast
+magick input.png -contrast-stretch 2%x1% output.png
+
+# Warm color temperature
+magick input.png -modulate 100,110,105 output.png
+
+# Cool color temperature
+magick input.png -modulate 100,90,95 output.png
+
+# Desaturate (muted colors)
+magick input.png -modulate 100,70,100 output.png
+
+# Convert to grayscale
+magick input.png -colorspace Gray output.png
+
+# Sepia tone
+magick input.png -sepia-tone 80% output.png
+```
+
+### Compositing
+
+```bash
+# Overlay watermark (bottom-right, 20% opacity)
+magick base.png watermark.png -gravity southeast -geometry +20+20 \
+ -compose dissolve -define compose:args=20 -composite output.png
+
+# Side-by-side comparison
+magick input1.png input2.png +append comparison.png
+
+# Vertical stack
+magick input1.png input2.png -append stack.png
+
+# Add padding/border
+magick input.png -bordercolor white -border 40 output.png
+
+# Add rounded corners
+magick input.png \( +clone -alpha extract -draw \
+ "roundrectangle 0,0,%[fx:w-1],%[fx:h-1],20,20" \) \
+ -alpha off -compose CopyOpacity -composite rounded.png
+```
+
+### Batch Processing
+
+```bash
+# Resize all PNGs in directory
+for f in ~/Documents/nanobanana_generated/*.png; do
+ magick "$f" -resize 800x800 "${f%.png}_thumb.png"
+done
+
+# Convert all to WebP
+for f in ~/Documents/nanobanana_generated/*.png; do
+ magick "$f" -quality 85 "${f%.png}.webp"
+done
+```
+
+## Animation (GIF/Video from Multiple Frames)
+
+```bash
+# Create GIF from multiple images
+magick -delay 100 frame1.png frame2.png frame3.png animation.gif
+
+# Create MP4 from image sequence
+ffmpeg -framerate 1 -pattern_type glob -i '*.png' \
+ -c:v libx264 -pix_fmt yuv420p slideshow.mp4
+```
+
+## Note on 4K Output
+
+With Gemini 3.1 Flash's `imageSize: "4K"` option (up to 4096×4096), many traditional
+upscaling post-processing steps are no longer necessary. If your target platform accepts
+images at or below 4K resolution, generate at native 4K instead of generating at 1K
+and upscaling. This produces better detail and avoids upscaling artifacts.
+
+## Green Screen Transparency Pipeline
+
+Gemini cannot generate transparent backgrounds. Use this workaround:
+
+### 1. Generate with green screen prompt
+
+Append to any prompt:
+```
+on a solid bright green (#00FF00) chroma key background
+with a thin white outline separating the subject from the background
+```
+
+### 2. Remove green screen (ImageMagick)
+
+```bash
+magick input.png -fuzz 20% -transparent "#00FF00" output.png
+```
+
+### 3. Clean edges + trim (ImageMagick)
+
+```bash
+magick output.png -channel A -blur 0x1 -level 50%,100% -trim +repage final.png
+```
+
+### 4. Alternative (FFmpeg, better for batch)
+
+```bash
+ffmpeg -i input.png -vf "colorkey=0x00FF00:0.3:0.1,despill=type=green" -pix_fmt rgba output.png
+```
+
+### Tips
+- `-fuzz 20%` handles slight color variations at edges; increase to 25% for softer edges
+- The white outline in the prompt helps prevent color spill on subject edges
+- For batch processing, the FFmpeg approach is faster and handles despill automatically
+- Always verify edges after conversion — may need manual touchup for hair/fur
+
+## Quality Assessment
+
+```bash
+# Get image dimensions and info
+magick identify -verbose input.png | head -20
+
+# Check file size
+ls -lh input.png
+
+# Get exact pixel dimensions
+magick identify -format "%wx%h" input.png
+```
diff --git a/extensions/banana/references/presets.md b/extensions/banana/references/presets.md
new file mode 100644
index 0000000..2b6df0b
--- /dev/null
+++ b/extensions/banana/references/presets.md
@@ -0,0 +1,69 @@
+# Brand/Style Presets Reference
+
+> Load this on-demand when the user asks about presets or brand consistency.
+
+## Preset Schema
+
+Each preset is stored as `~/.banana/presets/NAME.json`:
+
+```json
+{
+ "name": "tech-saas",
+ "description": "Clean tech SaaS brand",
+ "colors": ["#2563EB", "#1E40AF", "#F8FAFC"],
+ "style": "clean minimal tech illustration, flat vectors, soft shadows",
+ "typography": "bold geometric sans-serif",
+ "lighting": "bright diffused studio, no harsh shadows",
+ "mood": "professional, trustworthy, modern",
+ "default_ratio": "16:9",
+ "default_resolution": "2K"
+}
+```
+
+## Example Presets
+
+### tech-saas
+- **Colors:** #2563EB, #1E40AF, #F8FAFC (blue + white)
+- **Style:** Clean minimal tech illustration, flat vectors, soft shadows
+- **Typography:** Bold geometric sans-serif
+- **Mood:** Professional, trustworthy, modern
+
+### luxury-brand
+- **Colors:** #1A1A1A, #C9A96E, #FAFAF5 (black + gold + cream)
+- **Style:** Elegant high-end photography, rich textures, deep contrast
+- **Typography:** Thin elegant serif, generous letter-spacing
+- **Mood:** Exclusive, sophisticated, aspirational
+
+### editorial-magazine
+- **Colors:** #000000, #FFFFFF, #FF3B30 (black + white + accent red)
+- **Style:** Bold editorial photography, strong geometric composition
+- **Typography:** Condensed all-caps sans-serif headlines
+- **Mood:** Bold, provocative, contemporary
+
+## How Presets Merge into Reasoning Brief
+
+When a preset is active, Claude uses its values as defaults for the Reasoning Brief:
+1. **Colors** → inform palette descriptions in Context and Style components
+2. **Style** → becomes the base for the Style component
+3. **Typography** → used for any text rendering
+4. **Lighting** → becomes the base for the Lighting component
+5. **Mood** → influences Action and Context components
+
+User instructions always override preset values. If a user says "make it dark"
+but the preset has bright lighting, follow the user's instruction.
+
+## Managing Presets
+
+```bash
+# List presets
+presets.py list
+
+# Show details
+presets.py show tech-saas
+
+# Create interactively (Claude fills in details from conversation)
+presets.py create NAME --colors "#hex,#hex" --style "..." --mood "..."
+
+# Delete
+presets.py delete NAME --confirm
+```
diff --git a/extensions/banana/references/prompt-engineering.md b/extensions/banana/references/prompt-engineering.md
new file mode 100644
index 0000000..29f8336
--- /dev/null
+++ b/extensions/banana/references/prompt-engineering.md
@@ -0,0 +1,411 @@
+# Prompt Engineering Reference — Claude Banana
+
+> Load this on-demand when constructing complex prompts or when the user
+> asks about prompt techniques. Do NOT load at startup.
+>
+> Aligned with Google's March 2026 "Ultimate Prompting Guide" for Gemini image generation.
+
+## The 6-Component Reasoning Brief
+
+Every image prompt should contain these components, written as natural
+narrative paragraphs — NEVER as comma-separated keyword lists.
+
+### 1. Subject
+The main focus of the image. Describe with physical specificity.
+
+**Good:** "A weathered Japanese ceramicist in his 70s, deep sun-etched
+wrinkles mapping decades of kiln work, calloused hands cradling a
+freshly thrown tea bowl with an irregular, organic rim"
+
+**Bad:** "old man, ceramic, bowl"
+
+### 2. Action
+What is happening. Movement, pose, gesture, state of being.
+
+**Good:** "leaning forward with intense concentration, gently smoothing
+the rim with a wet thumb, a thin trail of slip running down his wrist"
+
+**Bad:** "making pottery"
+
+### 3. Context
+Environment, setting, temporal and spatial details.
+
+**Good:** "inside a traditional wood-fired anagama kiln workshop,
+stacked shelves of drying pots visible in the soft background, late
+afternoon light filtering through rice paper screens"
+
+**Bad:** "workshop, afternoon"
+
+### 4. Composition
+Camera angle, shot type, framing, spatial relationships.
+
+**Good:** "intimate close-up shot from slightly below eye level,
+shallow depth of field isolating the hands and bowl against the
+soft bokeh of the workshop behind"
+
+**Bad:** "close up"
+
+### 5. Lighting
+Light source, quality, direction, temperature, shadows.
+
+**Good:** "warm directional light from a single high window camera-left,
+creating gentle Rembrandt lighting on the face with a soft triangle
+of light on the shadow-side cheek, deep warm shadows in the workshop"
+
+**Bad:** "natural lighting"
+
+### 6. Style
+Art medium, aesthetic reference, technical photographic details.
+
+**Good:** "captured with a Sony A7R IV, 85mm f/1.4 GM lens, Kodak Portra
+400 color grading with lifted shadows and muted earth tones, reminiscent
+of Dorothea Lange's documentary portraiture"
+
+**Bad:** "photorealistic, 8K, masterpiece"
+
+## Domain Mode Modifier Libraries
+
+### Cinema Mode
+**Camera specs:** RED V-Raptor, ARRI Alexa 65, Sony Venice 2, Blackmagic URSA
+**Lenses:** Cooke S7/i, Zeiss Supreme Prime, Atlas Orion anamorphic
+**Film stocks:** Kodak Vision3 500T (tungsten), Kodak Vision3 250D (daylight), Fuji Eterna Vivid
+**Lighting setups:** three-point, chiaroscuro, Rembrandt, split, butterfly, rim/backlight
+**Shot types:** establishing wide, medium close-up, extreme close-up, Dutch angle, overhead crane, Steadicam tracking
+**Color grading:** teal and orange, desaturated cold, warm vintage, high-contrast noir
+
+### Product Mode
+**Surfaces:** polished marble, brushed concrete, raw linen, acrylic riser, gradient sweep
+**Lighting:** softbox diffused, hard key with fill card, rim separation, tent lighting, light painting
+**Angles:** 45-degree hero, flat lay, three-quarter, straight-on, worm's-eye
+**Style refs:** Apple product photography, Aesop minimal, Bang & Olufsen clean, luxury cosmetics
+
+### Portrait Mode
+**Focal lengths:** 85mm (classic), 105mm (compression), 135mm (telephoto), 50mm (environmental)
+**Apertures:** f/1.4 (dreamy bokeh), f/2.8 (subject-sharp), f/5.6 (environmental context)
+**Pose language:** candid mid-gesture, direct-to-camera confrontational, profile silhouette, over-shoulder glance
+**Skin/texture:** freckles visible, pores at macro distance, catch light in eyes, subsurface scattering
+
+### Editorial/Fashion Mode
+**Publication refs:** Vogue Italia, Harper's Bazaar, GQ, National Geographic, Kinfolk
+**Styling notes:** layered textures, statement accessories, monochromatic palette, contrast patterns
+**Locations:** marble staircase, rooftop at golden hour, industrial loft, desert dunes, neon-lit alley
+**Poses:** power stance, relaxed editorial lean, movement blur, fabric in wind
+
+### UI/Web Mode
+**Styles:** flat vector, isometric 3D, line art, glassmorphism, neumorphism, material design
+**Colors:** specify exact hex or descriptive palette (e.g., "cool blues #2563EB to #1E40AF")
+**Sizing:** design at 2x for retina, specify exact pixel dimensions needed
+**Backgrounds:** transparent (request solid white then post-process), gradient, solid color
+
+### Logo Mode
+**Construction:** geometric primitives, golden ratio, grid-based, negative space
+**Typography:** bold sans-serif, elegant serif, custom lettermark, monogram
+**Colors:** max 2-3 colors, works in monochrome, high contrast
+**Output:** request on solid white background, post-process to transparent
+
+### Landscape Mode
+**Depth layers:** foreground interest, midground subject, background atmosphere
+**Atmospherics:** fog, mist, haze, volumetric light rays, dust particles
+**Time of day:** blue hour (pre-dawn), golden hour, magic hour (post-sunset), midnight blue
+**Weather:** dramatic storm clouds, clearing after rain, snow-covered, sun-dappled
+
+### Infographic Mode
+**Layout:** modular sections, clear visual hierarchy, bento grid, flow top-to-bottom
+**Text:** use quotes for exact text, descriptive font style, specify size hierarchy
+**Data viz:** bar charts, pie charts, flow diagrams, timelines, comparison tables
+**Colors:** high-contrast, accessible palette, consistent brand colors
+
+### Abstract Mode
+**Geometry:** fractals, voronoi tessellation, spirals, fibonacci, organic flow, crystalline
+**Textures:** marble veining, fluid dynamics, smoke wisps, ink diffusion, watercolor bleed
+**Color palettes:** analogous harmony, complementary clash, monochromatic gradient, neon-on-black
+**Styles:** generative art, data visualization art, glitch, procedural, macro photography of materials
+
+## Advanced Techniques
+
+### Character Consistency (Multi-turn)
+Use `gemini_chat` and maintain descriptive anchors:
+- First turn: Generate character with exhaustive physical description
+- Following turns: Reference "the same character" + repeat 2-3 key identifiers
+- Key identifiers: hair color/style, distinctive clothing, facial feature
+
+**Multi-image reference technique** (3.1 Flash):
+- Provide up to 4-5 character reference images in the conversation
+- Assign distinct names to each character ("Character A: the red-haired knight")
+- Model preserves features across different angles, actions, and environments
+- Works best when reference images show the character from multiple angles
+
+### Style Transfer Without Reference Images
+Describe the target style exhaustively instead of referencing an image:
+```
+Render this scene in the style of a 1950s travel poster: flat areas of
+color in a limited palette of teal, coral, and cream. Bold geometric
+shapes with visible paper texture. Hand-lettered title text with a
+mid-century modern typeface feel.
+```
+
+### Text Rendering Tips
+- Quote exact text: `with the text "OPEN DAILY" in bold condensed sans-serif`
+- **25 characters or less** — this is the practical limit for reliable rendering
+- **2-3 distinct phrases max** — more text fragments degrade quality
+- Describe font characteristics, not font names
+- Specify placement: "centered at the top third", "along the bottom edge"
+- High contrast: light text on dark, or vice versa
+- **Text-first hack:** Establish the text concept conversationally first ("I need a sign that says FRESH BREAD"), then generate — the model anchors on text mentioned early
+- Expect creative font interpretations, not exact replication of described styles
+
+### Positive Framing (No Negative Prompts)
+Gemini does NOT support negative prompts. Rephrase exclusions:
+- Instead of "no blur" → "sharp, in-focus, tack-sharp detail"
+- Instead of "no people" → "empty, deserted, uninhabited"
+- Instead of "no text" → "clean, uncluttered, text-free"
+- Instead of "not dark" → "brightly lit, high-key lighting"
+
+### Search-Grounded Generation
+For images based on real-world data (weather, events, statistics),
+Gemini can use Google Search grounding to incorporate live information.
+Useful for infographics with current data.
+
+**Three-part formula for search-grounded prompts:**
+1. `[Source/Search request]` — What to look up
+2. `[Analytical task]` — What to analyze or extract
+3. `[Visual translation]` — How to render it as an image
+
+**Example:** "Search for the current top 5 programming languages by GitHub usage in 2026, analyze their relative popularity percentages, then generate a clean infographic bar chart with the language logos and percentages in a modern dark theme."
+
+## Prompt Adaptation Rules
+
+When adapting prompts from the claude-prompts database (Midjourney/DALL-E/etc.)
+to Gemini's natural language format:
+
+| Source Syntax | Gemini Equivalent |
+|---------------|-------------------|
+| `--ar 16:9` | Call `set_aspect_ratio("16:9")` separately |
+| `--v 6`, `--style raw` | Remove — Gemini has no version/style flags |
+| `--chaos 50` | Describe variety: "unexpected, surreal composition" |
+| `--no trees` | Positive framing: "open clearing with no vegetation" |
+| `(word:1.5)` weight | Descriptive emphasis: "prominently featuring [word]" |
+| `8K, masterpiece, ultra-detailed` | Keep only "ultra-realistic, high resolution"; remove the rest |
+| Comma-separated tags | Expand into descriptive narrative paragraphs |
+| `shot on Hasselblad` | Keep — camera specs work well in Gemini |
+
+## Common Prompt Mistakes
+
+1. **Keyword stuffing** — stacking generic quality terms ("8K, masterpiece, best quality") adds nothing. Use only "ultra-realistic, high resolution" at the end
+2. **Tag lists** — Gemini wants prose, not "red car, sunset, mountain, cinematic"
+3. **Missing lighting** — The single biggest quality differentiator
+4. **No composition direction** — Results in generic centered framing
+5. **Vague style** — "make it look cool" vs specific art direction
+6. **Ignoring aspect ratio** — Always set before generating
+7. **Overlong prompts** — Diminishing returns past ~200 words; be precise, not verbose
+8. **Text longer than ~25 characters** — Rendering degrades rapidly past this limit
+9. **Burying key details at the end** — In long prompts, details placed last may be deprioritized; put critical specifics (exact text, key constraints) in the first third of the prompt
+10. **Not iterating with follow-up prompts** — Use `gemini_chat` for progressive refinement instead of trying to get everything right in one generation
+
+## Proven Prompt Templates
+
+> Extracted from 2,500+ tested prompts. These patterns consistently produce
+> high-quality results. Use them as starting points and adapt to the request.
+
+### The Winning Formula (Weight Distribution)
+
+| Component | Weight | What to include |
+|-----------|--------|-----------------|
+| **Subject** | 40% | Age, skin tone, hair color/style, eye color, body type, expression |
+| **Styling** | 25% | Brand names, textures, fit, accessories, colors |
+| **Environment** | 15% | Location + time of day + context details |
+| **Camera** | 10% | Camera model + lens + focal length + shot type |
+| **Lighting** | 10% | Quality, direction, color temperature, shadows |
+
+### Instagram Ad / Social Media
+
+**Pattern:** `[Subject with age/appearance] + [outfit with brand/texture] + [action verb] + [setting] + [camera spec] + [lighting] + [platform aesthetic]`
+
+**Example (Product Placement):**
+```
+Hyper-realistic gym selfie of athletic 24yo influencer with glowing olive
+skin, wearing crinkle-textured athleisure set in mauve. iPhone 16 Pro Max
+front-facing portrait mode capturing sweat droplets on collarbones, hazel
+eyes enhanced by gym LED lighting. Mirror reflection shows perfect form,
+golden morning light through floor-to-ceiling windows. Frayed chestnut
+ponytail with baby hairs, ultra-detailed skin texture, trending aesthetic.
+```
+
+**Example (Lifestyle Ad):**
+```
+A 24-year-old blonde fitness model in a high-energy sports drink
+advertisement. Mid-run on a beach, wearing a vibrant orange sports bra
+and black shorts, playful smile and sparkling blue eyes exuding vitality.
+Bottle of the drink held in hand, waves crashing in background. Shot on
+Nikon D850 with 70-200mm f/2.8 lens, natural light, fast shutter speed
+capturing motion. Ultra-realistic skin texture, water droplets, product
+label clearly visible.
+```
+
+**Example (Luxury Lifestyle):**
+```
+Gorgeous Instagram model wearing a designer silk gown, luxury rooftop
+restaurant, golden hour lighting, champagne in hand, luxurious aspirational
+lifestyle. Captured with Sony A7R IV, 85mm f/1.4 lens, shallow depth of
+field, warm color grading.
+```
+
+### Product / Commercial Photography
+
+**Pattern:** `[Product with brand/detail] + [dynamic elements] + [surface/setting] + "commercial photography for advertising campaign" + [lighting] + "high resolution"`
+
+**Example (Beverage):**
+```
+Gatorade bottle with condensation dripping down the sides, surrounded by
+lightning bolts and a burst of vibrant blue and orange light rays. The
+Gatorade logo is prominently displayed on the bottle, with splashes of
+water frozen in mid-air. Commercial food photography for an advertising
+campaign, high resolution, high level of detail, vibrant complementary
+colors.
+```
+
+**Example (Food):**
+```
+In and Out burger with layers of fresh lettuce, melted cheese, and pretzel
+bun, placed on a white surface with the In and Out logo subtly glowing in
+the background. Falling french fries and golden light, warm scene.
+Commercial food photography for an advertising campaign, high resolution,
+high level of detail, vibrant complementary colors.
+```
+
+### Fashion / Editorial
+
+**Pattern:** `[Subject with ethnicity/age/features] + [outfit with texture/brand/cut] + [location] + [pose/action] + [camera + lens] + [lighting quality]`
+
+**Example (Street Style):**
+```
+A 24-year-old female AI influencer posing confidently in an urban cityscape
+during golden hour. Flawless sun-kissed skin, long wavy brown hair, deep
+green eyes. Wearing a chic streetwear outfit — oversized beige blazer,
+white top, high-waisted jeans. Captured with Sony A7R IV at 85mm f/1.4,
+shallow depth of field with warm golden bokeh.
+```
+
+**Example (High Fashion):**
+```
+Stunning 24-year-old woman, long platinum blonde hair, radiant skin,
+piercing blue eyes, dressed in a chic pastel blazer with a modern
+minimalist aesthetic, soft sunlight glow, high-end fashion appeal.
+Shot on Canon EOS R5, 85mm f/1.2 lens.
+```
+
+**Example (Avant-Garde):**
+```
+A blonde fitness model transformed into a runway-ready fashion icon,
+wearing a bold avant-garde outfit: cropped leather jacket with neon pink
+accents, paired with high-waisted athletic shorts and knee-high boots.
+Captured mid-stride on a minimalist white runway, playful twinkle in her
+eye, dramatic studio lighting from above.
+```
+
+### SaaS / Tech Marketing
+
+**Pattern:** `[UI mockup or abstract visual] + "on [dark/light] background" + [specific colors with hex] + [typography description] + "clean, premium SaaS aesthetic" + [glassmorphism/gradient/glow effects]`
+
+**Example (Dashboard Hero):**
+```
+A floating glassmorphism UI card on a deep charcoal background showing a
+content analytics dashboard with a rising line graph in teal (#14B8A6),
+bar charts in coral (#F97316), and a circular progress indicator at 94%.
+Subtle grid lines, frosted glass effect with 20% opacity, teal glow
+bleeding from the card edges. Clean premium SaaS aesthetic, no text
+smaller than headline size.
+```
+
+**Example (Feature Highlight):**
+```
+An isometric 3D illustration of interconnected data nodes on a dark navy
+background. Each node is a glowing teal sphere connected by thin luminous
+lines, forming a constellation pattern. One central node pulses brighter
+with radiating rings. Modern tech illustration style with subtle depth
+of field, volumetric lighting from below.
+```
+
+**Example (Comparison/Before-After):**
+```
+Split-screen image: left side shows a cluttered, dim workspace with
+scattered papers, red error indicators, and a frustrated expression
+conveyed through a cracked coffee mug and tangled cables. Right side
+shows a clean, organized dashboard interface glowing in teal and white
+on a dark background, with smooth flowing lines and checkmarks. A sharp
+vertical dividing line separates chaos from clarity.
+```
+
+### Logo / Branding
+
+**Pattern:** `[Product/bottle/item] + "with [brand element] prominently displayed" + [dynamic visual elements] + "commercial photography" + [lighting style] + "high resolution, vibrant complementary colors"`
+
+**Example:**
+```
+A sleek matte black bottle with a minimal white logo mark centered on the
+label, surrounded by swirling gradient ribbons of teal and coral light.
+The bottle sits on a reflective dark surface, sharp studio rim lighting
+separating it from the background. Product photography for luxury
+branding, high resolution, dramatic contrast.
+```
+
+### Key Tactics That Make Prompts Work
+
+1. **Name real cameras** — "Sony A7R IV", "Canon EOS R5", "iPhone 16 Pro Max" anchor realism
+2. **Specify exact lens** — "85mm f/1.4" gives the model precise depth-of-field information
+3. **Use age + ethnicity + features** — "24yo with olive skin, hazel eyes" beats "a person"
+4. **Name brands for styling** — "Lululemon mat", "Tom Ford suit" triggers specific visual associations
+5. **Include micro-details** — "sweat droplets on collarbones", "baby hairs stuck to neck"
+6. **Add platform context** — "Instagram aesthetic", "commercial photography for advertising"
+7. **Describe textures** — "crinkle-textured", "metallic silver", "frosted glass"
+8. **Use action verbs** — "mid-run", "posing confidently", "captured mid-stride"
+9. **End with "ultra-realistic, high resolution"** — these two specific anchors help on Gemini. Avoid generic stacking like "8K, masterpiece, best quality" which adds no value
+10. **For products, say "prominently displayed"** — ensures the product/logo isn't hidden
+
+### Anti-Patterns (What NOT to Do)
+
+- **"A dark-themed Instagram ad showing..."** — too meta, describes the concept not the image
+- **"A sleek SaaS dashboard visualization..."** — abstract, no visual anchors
+- **"Modern, clean, professional..."** — vague adjectives that mean nothing to the model
+- **"A bold call to action with..."** — describes marketing intent, not visual content
+- **Describing what the viewer should feel** — instead, describe what creates that feeling
+
+## Safety Filter Rephrase Strategies
+
+Gemini's safety filters (Layer 2: server-side output filter) cannot be disabled.
+When a prompt is blocked, the only path forward is rephrasing.
+
+### Common Trigger Categories
+
+| Category | Triggers on | Rephrase approach |
+|----------|------------|-------------------|
+| Violence/weapons | Combat, blood, injuries, firearms | Use metaphor or aftermath: "battle-worn" → "weathered veteran" |
+| Medical/gore | Surgery, wounds, anatomical detail | Abstract or clinical: "open wound" → "medical illustration" |
+| Real public figures | Named celebrities, politicians | Use archetypes: "Elon Musk" → "a tech entrepreneur in a minimalist office" |
+| Children + risk | Minors in any ambiguous context | Add safety context: specify educational, family, or playful framing |
+| NSFW/suggestive | Revealing clothing, intimate poses | Use artistic framing: "fashion editorial, fully clothed, editorial pose" |
+
+### Rephrase Patterns
+
+1. **Abstraction** — Replace specific dangerous elements with abstract concepts
+2. **Artistic framing** — Frame content as art, editorial, or documentary
+3. **Metaphor** — Use symbolic language instead of literal descriptions
+4. **Positive emphasis** — Describe what IS present, not what's dangerous
+5. **Context shift** — Move from threatening to educational/professional context
+
+### Example Rephrases
+
+| Blocked prompt | Successful rephrase |
+|----------------|---------------------|
+| "a soldier in combat firing a rifle" | "a determined soldier standing guard at dawn, rifle slung over shoulder, morning mist over the outpost" |
+| "a scary horror monster" | "a fantastical creature from a dark fairy tale, intricate organic textures, bioluminescent accents, concept art style" |
+| "dog in a fight" | "a friendly golden retriever playing energetically in a sunny park, action shot, joyful expression" |
+| "medical surgery scene" | "a clean modern operating room viewed from the observation gallery, soft blue surgical lights, professional documentary style" |
+| "celebrity portrait of [name]" | "a distinguished middle-aged man in a tailored navy suit, warm studio lighting, editorial portrait style" |
+
+### Key Principle
+
+Layer 2 (output filter) analyzes the generated image, not just the prompt.
+Even well-phrased prompts can be blocked if the model's interpretation triggers
+the output filter. When this happens, try shifting the visual concept further
+from the trigger rather than just changing words.
diff --git a/extensions/banana/references/seo-image-presets.md b/extensions/banana/references/seo-image-presets.md
new file mode 100644
index 0000000..88d428f
--- /dev/null
+++ b/extensions/banana/references/seo-image-presets.md
@@ -0,0 +1,137 @@
+# SEO Image Presets
+
+Pre-configured presets for common SEO image use cases. These map to banana's
+preset format (see `references/presets.md` for schema details).
+
+## Preset Templates
+
+### og-default — Standard OG/Social Preview
+
+```json
+{
+ "name": "og-default",
+ "description": "Clean, professional OG image for social sharing",
+ "aspect_ratio": "16:9",
+ "resolution": "1K",
+ "domain_mode": "Product",
+ "style": {
+ "colors": ["#FFFFFF", "#F5F5F5"],
+ "mood": "Professional, clean, trustworthy",
+ "lighting": "Bright, even studio lighting with soft shadows",
+ "typography": "Modern sans-serif if text needed"
+ },
+ "post_processing": "magick output.png -resize 1200x630^ -gravity center -extent 1200x630 output-og.webp"
+}
+```
+
+### blog-hero — Widescreen Blog Hero Image
+
+```json
+{
+ "name": "blog-hero",
+ "description": "Dramatic widescreen hero for blog posts",
+ "aspect_ratio": "16:9",
+ "resolution": "2K",
+ "domain_mode": "Cinema",
+ "style": {
+ "colors": ["contextual"],
+ "mood": "Dramatic, atmospheric, editorial",
+ "lighting": "Golden hour or moody blue hour, directional",
+ "typography": "None — image only"
+ },
+ "post_processing": "magick output.png -quality 85 output-hero.webp"
+}
+```
+
+### product-white — E-commerce Product Shot
+
+```json
+{
+ "name": "product-white",
+ "description": "Clean white background product photography",
+ "aspect_ratio": "4:3",
+ "resolution": "2K",
+ "domain_mode": "Product",
+ "style": {
+ "colors": ["#FFFFFF"],
+ "mood": "Clean, professional, catalog-ready",
+ "lighting": "360-degree soft studio lighting, minimal shadows",
+ "typography": "None"
+ },
+ "prompt_suffix": "Studio product photography, clean white background, professional catalog shot, high resolution",
+ "post_processing": "magick output.png -fuzz 5% -transparent white output-transparent.png"
+}
+```
+
+### social-square — Social Media Square
+
+```json
+{
+ "name": "social-square",
+ "description": "1:1 square image for social media platforms",
+ "aspect_ratio": "1:1",
+ "resolution": "1K",
+ "domain_mode": "UI/Web",
+ "style": {
+ "colors": ["brand-contextual"],
+ "mood": "Engaging, scroll-stopping, platform-native",
+ "lighting": "Bright, even, high-contrast",
+ "typography": "Bold sans-serif if text needed, under 25 characters"
+ }
+}
+```
+
+### infographic-vertical — Data-Heavy Infographic
+
+```json
+{
+ "name": "infographic-vertical",
+ "description": "Tall vertical infographic for data visualization",
+ "aspect_ratio": "2:3",
+ "resolution": "4K",
+ "domain_mode": "Infographic",
+ "style": {
+ "colors": ["brand-contextual", "data-visualization palette"],
+ "mood": "Informative, structured, authoritative",
+ "lighting": "Flat, even, no dramatic shadows",
+ "typography": "Clear hierarchy — headline, subheads, body, captions"
+ },
+ "notes": "Use thinking: high for better text rendering accuracy"
+}
+```
+
+### favicon-mark — Favicon / App Icon
+
+```json
+{
+ "name": "favicon-mark",
+ "description": "Minimal iconic mark for favicon or app icon",
+ "aspect_ratio": "1:1",
+ "resolution": "512",
+ "domain_mode": "Logo",
+ "style": {
+ "colors": ["2-3 brand colors max"],
+ "mood": "Minimal, recognizable, scalable",
+ "lighting": "Flat, no shadows",
+ "typography": "Single letter or symbol only"
+ },
+ "post_processing": "magick output.png -resize 32x32 favicon.ico && magick output.png -resize 180x180 apple-touch-icon.png"
+}
+```
+
+## Creating Custom Presets
+
+Users can create their own presets:
+```bash
+python3 ~/.claude/skills/seo-image-gen/scripts/presets.py create my-brand
+```
+
+This creates `~/.banana/presets/my-brand.json` with the full schema.
+Custom presets override SEO defaults when specified.
+
+## Preset Selection Logic
+
+1. If user specifies a use case command (og, hero, product), load the matching preset
+2. If user mentions a brand preset name, load from `~/.banana/presets/`
+3. Brand presets override SEO presets for colors, mood, and typography
+4. SEO presets always provide aspect ratio and resolution defaults
diff --git a/extensions/banana/scripts/batch.py b/extensions/banana/scripts/batch.py
new file mode 100755
index 0000000..b0640f2
--- /dev/null
+++ b/extensions/banana/scripts/batch.py
@@ -0,0 +1,97 @@
+#!/usr/bin/env python3
+"""Claude Banana — CSV Batch Workflow
+
+Parse a CSV file of image generation requests and output a structured plan.
+Claude then executes each row via MCP.
+
+Usage:
+ batch.py --csv path/to/file.csv
+
+CSV columns:
+ prompt (required), ratio, resolution, model, preset (all optional)
+
+Example CSV:
+ prompt,ratio,resolution
+ "coffee shop hero image",16:9,2K
+ "team photo placeholder",1:1,1K
+ "product shot on marble",4:3,2K
+"""
+
+import argparse
+import csv
+import json
+import sys
+from pathlib import Path
+
+# Inline pricing for estimates
+PRICING = {
+ "gemini-3.1-flash-image-preview": {"512": 0.020, "1K": 0.039, "2K": 0.078, "4K": 0.156},
+ "gemini-2.5-flash-image": {"512": 0.020, "1K": 0.039},
+}
+DEFAULT_MODEL = "gemini-3.1-flash-image-preview"
+DEFAULT_RESOLUTION = "1K"
+DEFAULT_RATIO = "1:1"
+
+
+def estimate_cost(model, resolution):
+ """Estimate cost for a single image."""
+ model_pricing = PRICING.get(model, PRICING[DEFAULT_MODEL])
+ return model_pricing.get(resolution, model_pricing.get("1K", 0.039))
+
+
+def main():
+ parser = argparse.ArgumentParser(description="Parse CSV batch and output generation plan")
+ parser.add_argument("--csv", required=True, help="Path to CSV file")
+ args = parser.parse_args()
+
+ csv_path = Path(args.csv).resolve()
+ if not csv_path.exists():
+ print(json.dumps({"error": True, "message": f"CSV not found: {csv_path}"}))
+ sys.exit(1)
+
+ rows = []
+ errors = []
+
+ try:
+ with open(csv_path, "r", newline="") as f:
+ reader = csv.DictReader(f)
+ if not reader.fieldnames or "prompt" not in reader.fieldnames:
+ print(json.dumps({"error": True, "message": "CSV must have a 'prompt' column header"}))
+ sys.exit(1)
+ for i, row in enumerate(reader, start=2): # Line 2+ (1 is header)
+ prompt = row.get("prompt", "").strip()
+ if not prompt:
+ errors.append(f"Row {i}: missing prompt")
+ continue
+
+ rows.append({
+ "row": i,
+ "prompt": prompt,
+ "ratio": row.get("ratio", "").strip() or DEFAULT_RATIO,
+ "resolution": row.get("resolution", "").strip() or DEFAULT_RESOLUTION,
+ "model": row.get("model", "").strip() or DEFAULT_MODEL,
+ "preset": row.get("preset", "").strip() or None,
+ })
+ except (csv.Error, UnicodeDecodeError) as e:
+ print(json.dumps({"error": True, "message": f"Failed to parse CSV: {e}"}))
+ sys.exit(1)
+
+ if errors:
+ print("Validation errors:")
+ for e in errors:
+ print(f" - {e}")
+ if not rows:
+ sys.exit(1)
+ print()
+
+ # Cost estimate
+ total_cost = sum(estimate_cost(r["model"], r["resolution"]) for r in rows)
+
+ # Output structured JSON for Claude to consume
+ print(json.dumps({"rows": rows, "total_count": len(rows),
+ "estimated_cost": round(total_cost, 3),
+ "errors": errors}, indent=2))
+
+
+if __name__ == "__main__":
+ main()
diff --git a/extensions/banana/scripts/cost_tracker.py b/extensions/banana/scripts/cost_tracker.py
new file mode 100755
index 0000000..3bcf43e
--- /dev/null
+++ b/extensions/banana/scripts/cost_tracker.py
@@ -0,0 +1,191 @@
+#!/usr/bin/env python3
+"""Claude Banana — Cost Tracker
+
+Track image generation costs, view summaries, and estimate batch costs.
+
+Usage:
+ cost_tracker.py log --model MODEL --resolution RES --prompt "summary"
+ cost_tracker.py summary
+ cost_tracker.py today
+ cost_tracker.py estimate --model MODEL --resolution RES --count N
+ cost_tracker.py reset --confirm
+"""
+
+import argparse
+import json
+import sys
+from datetime import datetime, timezone
+from pathlib import Path
+
+LEDGER_PATH = Path.home() / ".banana" / "costs.json"
+
+# Cost per image in USD (approximate, based on ~1,290 output tokens)
+PRICING = {
+ "gemini-3.1-flash-image-preview": {
+ "512": 0.020,
+ "1K": 0.039,
+ "2K": 0.078,
+ "4K": 0.156,
+ },
+ "gemini-2.5-flash-image": {
+ "512": 0.020,
+ "1K": 0.039,
+ },
+}
+
+# Batch API gets 50% discount
+BATCH_DISCOUNT = 0.5
+
+
+def _load_ledger():
+ """Load the cost ledger from disk."""
+ if not LEDGER_PATH.exists():
+ return {"total_cost": 0.0, "total_images": 0, "entries": [], "daily": {}}
+ with open(LEDGER_PATH, "r") as f:
+ return json.load(f)
+
+
+def _save_ledger(ledger):
+ """Save the cost ledger to disk."""
+ LEDGER_PATH.parent.mkdir(parents=True, exist_ok=True)
+ with open(LEDGER_PATH, "w") as f:
+ json.dump(ledger, f, indent=2)
+
+
+def _lookup_cost(model, resolution, batch=False):
+ """Look up cost for a model+resolution combination."""
+ model_pricing = PRICING.get(model)
+ if not model_pricing:
+ # Try partial match
+ for key in PRICING:
+ if key in model or model in key:
+ model_pricing = PRICING[key]
+ break
+ if not model_pricing:
+ print(f"Warning: Unknown model '{model}', using 3.1 Flash pricing", file=sys.stderr)
+ model_pricing = PRICING["gemini-3.1-flash-image-preview"]
+
+ valid_resolutions = {"512", "1K", "2K", "4K"}
+ if resolution not in valid_resolutions:
+ print(f"Warning: Unknown resolution '{resolution}', using 1K pricing", file=sys.stderr)
+ cost = model_pricing.get(resolution, model_pricing.get("1K", 0.039))
+ if batch:
+ cost *= BATCH_DISCOUNT
+ return cost
+
+
+def cmd_log(args):
+ """Log a generation to the ledger."""
+ ledger = _load_ledger()
+ cost = _lookup_cost(args.model, args.resolution, getattr(args, "batch", False))
+ today = datetime.now(timezone.utc).strftime("%Y-%m-%d")
+ now = datetime.now(timezone.utc).strftime("%Y-%m-%dT%H:%M:%S")
+
+ entry = {
+ "ts": now,
+ "model": args.model,
+ "res": args.resolution,
+ "cost": cost,
+ "prompt": args.prompt[:100],
+ }
+
+ ledger["entries"].append(entry)
+ ledger["total_cost"] = round(ledger["total_cost"] + cost, 4)
+ ledger["total_images"] += 1
+
+ if today not in ledger["daily"]:
+ ledger["daily"][today] = {"count": 0, "cost": 0.0}
+ ledger["daily"][today]["count"] += 1
+ ledger["daily"][today]["cost"] = round(ledger["daily"][today]["cost"] + cost, 4)
+
+ _save_ledger(ledger)
+ print(json.dumps({"logged": True, "cost": cost, "total_cost": ledger["total_cost"],
+ "total_images": ledger["total_images"]}))
+
+
+def cmd_summary(args):
+ """Show cost summary."""
+ ledger = _load_ledger()
+ print(f"Total images: {ledger['total_images']}")
+ print(f"Total cost: ${ledger['total_cost']:.3f}")
+ print()
+
+ daily = ledger.get("daily", {})
+ if daily:
+ # Show last 7 days
+ sorted_days = sorted(daily.keys(), reverse=True)[:7]
+ print("Last 7 days:")
+ for day in sorted_days:
+ d = daily[day]
+ print(f" {day}: {d['count']} images, ${d['cost']:.3f}")
+ else:
+ print("No usage recorded yet.")
+
+
+def cmd_today(args):
+ """Show today's usage."""
+ ledger = _load_ledger()
+ today = datetime.now(timezone.utc).strftime("%Y-%m-%d")
+ daily = ledger.get("daily", {}).get(today, {"count": 0, "cost": 0.0})
+ print(f"Today ({today}): {daily['count']} images, ${daily['cost']:.3f}")
+
+
+def cmd_estimate(args):
+ """Estimate cost for a batch."""
+ cost_per = _lookup_cost(args.model, args.resolution, getattr(args, "batch", False))
+ total = round(cost_per * args.count, 3)
+ print(f"Model: {args.model}")
+ print(f"Resolution: {args.resolution}")
+ print(f"Count: {args.count}")
+ print(f"Cost/image: ${cost_per:.3f}")
+ print(f"Total est: ${total:.3f}")
+ if not getattr(args, "batch", False):
+ batch_total = round(cost_per * BATCH_DISCOUNT * args.count, 3)
+ print(f"Batch est: ${batch_total:.3f} (50% discount)")
+
+
+def cmd_reset(args):
+ """Reset the ledger."""
+ if not args.confirm:
+ print("Error: Pass --confirm to reset the cost ledger.", file=sys.stderr)
+ sys.exit(1)
+ _save_ledger({"total_cost": 0.0, "total_images": 0, "entries": [], "daily": {}})
+ print("Cost ledger reset.")
+
+
+def main():
+ parser = argparse.ArgumentParser(description="Claude Banana Cost Tracker")
+ sub = parser.add_subparsers(dest="command", required=True)
+
+ # log
+ p_log = sub.add_parser("log", help="Log a generation")
+ p_log.add_argument("--model", required=True, help="Model ID")
+ p_log.add_argument("--resolution", required=True, help="Resolution (512, 1K, 2K, 4K)")
+ p_log.add_argument("--prompt", required=True, help="Brief prompt description")
+ p_log.add_argument("--batch", action="store_true", help="Batch API (50%% discount)")
+
+ # summary
+ sub.add_parser("summary", help="Show cost summary")
+
+ # today
+ sub.add_parser("today", help="Show today's usage")
+
+ # estimate
+ p_est = sub.add_parser("estimate", help="Estimate batch cost")
+ p_est.add_argument("--model", required=True, help="Model ID")
+ p_est.add_argument("--resolution", required=True, help="Resolution (512, 1K, 2K, 4K)")
+ p_est.add_argument("--count", required=True, type=int, help="Number of images")
+ p_est.add_argument("--batch", action="store_true", help="Use batch pricing (50%% discount)")
+
+ # reset
+ p_reset = sub.add_parser("reset", help="Reset cost ledger")
+ p_reset.add_argument("--confirm", action="store_true", help="Confirm reset")
+
+ args = parser.parse_args()
+ cmds = {"log": cmd_log, "summary": cmd_summary, "today": cmd_today,
+ "estimate": cmd_estimate, "reset": cmd_reset}
+ cmds[args.command](args)
+
+
+if __name__ == "__main__":
+ main()
diff --git a/extensions/banana/scripts/edit.py b/extensions/banana/scripts/edit.py
new file mode 100755
index 0000000..a3e6aae
--- /dev/null
+++ b/extensions/banana/scripts/edit.py
@@ -0,0 +1,141 @@
+#!/usr/bin/env python3
+"""Claude Banana — Direct API Fallback: Image Editing
+
+Edit images via Gemini REST API when MCP is unavailable.
+Uses only Python stdlib (no pip dependencies).
+
+Usage:
+ edit.py --image path/to/image.png --prompt "remove the background"
+ [--model MODEL] [--api-key KEY]
+"""
+
+import argparse
+import base64
+import json
+import os
+import sys
+import urllib.request
+from datetime import datetime
+from pathlib import Path
+
+DEFAULT_MODEL = "gemini-3.1-flash-image-preview"
+OUTPUT_DIR = Path.home() / "Documents" / "nanobanana_generated"
+API_BASE = "https://generativelanguage.googleapis.com/v1beta/models"
+
+
+def edit_image(image_path, prompt, model, api_key):
+ """Call Gemini API to edit an image."""
+ image_path = Path(image_path).resolve()
+ if not image_path.exists():
+ print(json.dumps({"error": True, "message": f"Image not found: {image_path}"}))
+ sys.exit(1)
+
+ # Read and encode image
+ with open(image_path, "rb") as f:
+ image_b64 = base64.b64encode(f.read()).decode("utf-8")
+
+ # Determine MIME type
+ suffix = image_path.suffix.lower()
+ mime_types = {".png": "image/png", ".jpg": "image/jpeg", ".jpeg": "image/jpeg",
+ ".webp": "image/webp", ".gif": "image/gif"}
+ mime_type = mime_types.get(suffix, "image/png")
+
+ url = f"{API_BASE}/{model}:generateContent?key={api_key}"
+
+ body = {
+ "contents": [
+ {
+ "parts": [
+ {"text": prompt},
+ {"inlineData": {"mimeType": mime_type, "data": image_b64}},
+ ]
+ }
+ ],
+ "generationConfig": {
+ "responseModalities": ["TEXT", "IMAGE"],
+ },
+ }
+
+ data = json.dumps(body).encode("utf-8")
+ req = urllib.request.Request(
+ url,
+ data=data,
+ headers={"Content-Type": "application/json"},
+ method="POST",
+ )
+
+ try:
+ with urllib.request.urlopen(req, timeout=120) as resp:
+ result = json.loads(resp.read().decode("utf-8"))
+ except urllib.error.HTTPError as e:
+ error_body = e.read().decode("utf-8") if e.fp else ""
+ print(json.dumps({"error": True, "status": e.code, "message": error_body}))
+ sys.exit(1)
+ except urllib.error.URLError as e:
+ print(json.dumps({"error": True, "message": str(e.reason)}))
+ sys.exit(1)
+
+ # Extract image from response
+ candidates = result.get("candidates", [])
+ if not candidates:
+ finish_reason = result.get("promptFeedback", {}).get("blockReason", "UNKNOWN")
+ print(json.dumps({"error": True, "message": f"No candidates returned. Reason: {finish_reason}"}))
+ sys.exit(1)
+
+ parts = candidates[0].get("content", {}).get("parts", [])
+ image_data = None
+ text_response = ""
+
+ for part in parts:
+ if "inlineData" in part:
+ image_data = part["inlineData"]["data"]
+ elif "text" in part:
+ text_response = part["text"]
+
+ if not image_data:
+ finish_reason = candidates[0].get("finishReason", "UNKNOWN")
+ print(json.dumps({"error": True, "message": f"No image in response. finishReason: {finish_reason}"}))
+ sys.exit(1)
+
+ # Save image
+ OUTPUT_DIR.mkdir(parents=True, exist_ok=True)
+ timestamp = datetime.now().strftime("%Y%m%d_%H%M%S_%f")
+ filename = f"banana_edit_{timestamp}.png"
+ output_path = (OUTPUT_DIR / filename).resolve()
+
+ with open(output_path, "wb") as f:
+ f.write(base64.b64decode(image_data))
+
+ return {
+ "path": str(output_path),
+ "model": model,
+ "source": str(image_path),
+ "text": text_response,
+ }
+
+
+def main():
+ parser = argparse.ArgumentParser(description="Edit images via Gemini REST API")
+ parser.add_argument("--image", required=True, help="Path to input image")
+ parser.add_argument("--prompt", required=True, help="Edit instruction")
+ parser.add_argument("--model", default=DEFAULT_MODEL, help=f"Model ID (default: {DEFAULT_MODEL})")
+ parser.add_argument("--api-key", default=None, help="Google AI API key (or set GOOGLE_AI_API_KEY env)")
+
+ args = parser.parse_args()
+
+ api_key = args.api_key or os.environ.get("GOOGLE_AI_API_KEY") or os.environ.get("GOOGLE_API_KEY")
+ if not api_key:
+ print(json.dumps({"error": True, "message": "No API key. Set GOOGLE_AI_API_KEY env or pass --api-key"}))
+ sys.exit(1)
+
+ result = edit_image(
+ image_path=args.image,
+ prompt=args.prompt,
+ model=args.model,
+ api_key=api_key,
+ )
+ print(json.dumps(result, indent=2))
+
+
+if __name__ == "__main__":
+ main()
diff --git a/extensions/banana/scripts/generate.py b/extensions/banana/scripts/generate.py
new file mode 100755
index 0000000..aa19a73
--- /dev/null
+++ b/extensions/banana/scripts/generate.py
@@ -0,0 +1,149 @@
+#!/usr/bin/env python3
+"""Claude Banana — Direct API Fallback: Image Generation
+
+Generate images via Gemini REST API when MCP is unavailable.
+Uses only Python stdlib (no pip dependencies).
+
+Usage:
+ generate.py --prompt "a cat in space" [--aspect-ratio 16:9] [--resolution 1K]
+ [--model MODEL] [--api-key KEY] [--thinking LEVEL] [--image-only]
+"""
+
+import argparse
+import base64
+import json
+import os
+import sys
+import urllib.request
+from datetime import datetime
+from pathlib import Path
+
+DEFAULT_MODEL = "gemini-3.1-flash-image-preview"
+DEFAULT_RESOLUTION = "1K"
+DEFAULT_RATIO = "1:1"
+OUTPUT_DIR = Path.home() / "Documents" / "nanobanana_generated"
+API_BASE = "https://generativelanguage.googleapis.com/v1beta/models"
+
+VALID_RATIOS = {"1:1", "16:9", "9:16", "4:3", "3:4", "2:3", "3:2",
+ "4:5", "5:4", "1:4", "4:1", "1:8", "8:1", "21:9"}
+VALID_RESOLUTIONS = {"512", "1K", "2K", "4K"}
+
+
+def generate_image(prompt, model, aspect_ratio, resolution, api_key,
+ thinking_level=None, image_only=False):
+ """Call Gemini API to generate an image."""
+ url = f"{API_BASE}/{model}:generateContent?key={api_key}"
+
+ modalities = ["IMAGE"] if image_only else ["TEXT", "IMAGE"]
+ body = {
+ "contents": [{"parts": [{"text": prompt}]}],
+ "generationConfig": {
+ "responseModalities": modalities,
+ "imageConfig": {
+ "aspectRatio": aspect_ratio,
+ "imageSize": resolution,
+ },
+ },
+ }
+
+ if thinking_level:
+ body["generationConfig"]["thinkingConfig"] = {"thinkingLevel": thinking_level}
+
+ data = json.dumps(body).encode("utf-8")
+ req = urllib.request.Request(
+ url,
+ data=data,
+ headers={"Content-Type": "application/json"},
+ method="POST",
+ )
+
+ try:
+ with urllib.request.urlopen(req, timeout=120) as resp:
+ result = json.loads(resp.read().decode("utf-8"))
+ except urllib.error.HTTPError as e:
+ error_body = e.read().decode("utf-8") if e.fp else ""
+ print(json.dumps({"error": True, "status": e.code, "message": error_body}))
+ sys.exit(1)
+ except urllib.error.URLError as e:
+ print(json.dumps({"error": True, "message": str(e.reason)}))
+ sys.exit(1)
+
+ # Extract image from response
+ candidates = result.get("candidates", [])
+ if not candidates:
+ finish_reason = result.get("promptFeedback", {}).get("blockReason", "UNKNOWN")
+ print(json.dumps({"error": True, "message": f"No candidates returned. Reason: {finish_reason}"}))
+ sys.exit(1)
+
+ parts = candidates[0].get("content", {}).get("parts", [])
+ image_data = None
+ text_response = ""
+
+ for part in parts:
+ if "inlineData" in part:
+ image_data = part["inlineData"]["data"]
+ elif "text" in part:
+ text_response = part["text"]
+
+ if not image_data:
+ finish_reason = candidates[0].get("finishReason", "UNKNOWN")
+ print(json.dumps({"error": True, "message": f"No image in response. finishReason: {finish_reason}"}))
+ sys.exit(1)
+
+ # Save image
+ OUTPUT_DIR.mkdir(parents=True, exist_ok=True)
+ timestamp = datetime.now().strftime("%Y%m%d_%H%M%S_%f")
+ filename = f"banana_{timestamp}.png"
+ output_path = (OUTPUT_DIR / filename).resolve()
+
+ with open(output_path, "wb") as f:
+ f.write(base64.b64decode(image_data))
+
+ return {
+ "path": str(output_path),
+ "model": model,
+ "aspect_ratio": aspect_ratio,
+ "resolution": resolution,
+ "text": text_response,
+ }
+
+
+def main():
+ parser = argparse.ArgumentParser(description="Generate images via Gemini REST API")
+ parser.add_argument("--prompt", required=True, help="Image generation prompt")
+ parser.add_argument("--aspect-ratio", default=DEFAULT_RATIO, help=f"Aspect ratio (default: {DEFAULT_RATIO})")
+ parser.add_argument("--resolution", default=DEFAULT_RESOLUTION, help=f"Resolution: 512, 1K, 2K, 4K (default: {DEFAULT_RESOLUTION})")
+ parser.add_argument("--model", default=DEFAULT_MODEL, help=f"Model ID (default: {DEFAULT_MODEL})")
+ parser.add_argument("--api-key", default=None, help="Google AI API key (or set GOOGLE_AI_API_KEY env)")
+ parser.add_argument("--thinking", default=None, choices=["minimal", "low", "medium", "high"], help="Thinking level")
+ parser.add_argument("--image-only", action="store_true", help="Return image only (no text)")
+
+ args = parser.parse_args()
+
+ if args.aspect_ratio not in VALID_RATIOS:
+ print(json.dumps({"error": True, "message": f"Invalid aspect ratio '{args.aspect_ratio}'. Valid: {sorted(VALID_RATIOS)}"}))
+ sys.exit(1)
+
+ if args.resolution not in VALID_RESOLUTIONS:
+ print(json.dumps({"error": True, "message": f"Invalid resolution '{args.resolution}'. Valid: {sorted(VALID_RESOLUTIONS)}"}))
+ sys.exit(1)
+
+ api_key = args.api_key or os.environ.get("GOOGLE_AI_API_KEY") or os.environ.get("GOOGLE_API_KEY")
+ if not api_key:
+ print(json.dumps({"error": True, "message": "No API key. Set GOOGLE_AI_API_KEY env or pass --api-key"}))
+ sys.exit(1)
+
+ result = generate_image(
+ prompt=args.prompt,
+ model=args.model,
+ aspect_ratio=args.aspect_ratio,
+ resolution=args.resolution,
+ api_key=api_key,
+ thinking_level=args.thinking,
+ image_only=args.image_only,
+ )
+ print(json.dumps(result, indent=2))
+
+
+if __name__ == "__main__":
+ main()
diff --git a/extensions/banana/scripts/presets.py b/extensions/banana/scripts/presets.py
new file mode 100755
index 0000000..7daffad
--- /dev/null
+++ b/extensions/banana/scripts/presets.py
@@ -0,0 +1,153 @@
+#!/usr/bin/env python3
+"""Claude Banana — Brand/Style Presets
+
+Manage reusable brand and style presets for consistent image generation.
+
+Usage:
+ presets.py list
+ presets.py show NAME
+ presets.py create NAME --colors "#hex,#hex" --style "..." [options]
+ presets.py delete NAME --confirm
+"""
+
+import argparse
+import json
+import re
+import sys
+from pathlib import Path
+
+PRESETS_DIR = Path.home() / ".banana" / "presets"
+
+
+def _ensure_dir():
+ """Ensure presets directory exists."""
+ PRESETS_DIR.mkdir(parents=True, exist_ok=True)
+
+
+def _sanitize_name(name):
+ """Sanitize preset name to prevent path traversal."""
+ # Strip path separators and keep only safe characters
+ safe = re.sub(r'[^a-zA-Z0-9_\-]', '', name)
+ if not safe:
+ print("Error: Preset name must contain only letters, numbers, hyphens, and underscores.", file=sys.stderr)
+ sys.exit(1)
+ return safe
+
+
+def _preset_path(name):
+ """Get path for a preset file."""
+ safe_name = _sanitize_name(name)
+ return PRESETS_DIR / f"{safe_name}.json"
+
+
+def _load_preset(name):
+ """Load a preset by name."""
+ path = _preset_path(name)
+ if not path.exists():
+ print(f"Error: Preset '{name}' not found.", file=sys.stderr)
+ sys.exit(1)
+ with open(path, "r") as f:
+ return json.load(f)
+
+
+def cmd_list(args):
+ """List available presets."""
+ _ensure_dir()
+ presets = sorted(PRESETS_DIR.glob("*.json"))
+ if not presets:
+ print("No presets found. Create one with: presets.py create NAME --style \"...\"")
+ return
+ print(f"Available presets ({len(presets)}):\n")
+ for p in presets:
+ try:
+ with open(p, "r") as f:
+ data = json.load(f)
+ desc = data.get("description", "No description")
+ print(f" {p.stem:20s} — {desc}")
+ except (json.JSONDecodeError, KeyError):
+ print(f" {p.stem:20s} — (invalid preset file)")
+
+
+def cmd_show(args):
+ """Show full preset details."""
+ preset = _load_preset(args.name)
+ print(json.dumps(preset, indent=2))
+
+
+def cmd_create(args):
+ """Create a new preset."""
+ _ensure_dir()
+ path = _preset_path(args.name)
+ if path.exists():
+ print(f"Error: Preset '{args.name}' already exists. Use a different name.", file=sys.stderr)
+ sys.exit(1)
+
+ colors = [c.strip() for c in args.colors.split(",")] if args.colors else []
+
+ preset = {
+ "name": args.name,
+ "description": args.description or f"Custom preset: {args.name}",
+ "colors": colors,
+ "style": args.style or "",
+ "typography": args.typography or "",
+ "lighting": args.lighting or "",
+ "mood": args.mood or "",
+ "default_ratio": args.ratio or "16:9",
+ "default_resolution": args.resolution or "2K",
+ }
+
+ with open(path, "w") as f:
+ json.dump(preset, f, indent=2)
+
+ print(f"Preset '{args.name}' created at {path}")
+ print(json.dumps(preset, indent=2))
+
+
+def cmd_delete(args):
+ """Delete a preset."""
+ if not args.confirm:
+ print("Error: Pass --confirm to delete the preset.", file=sys.stderr)
+ sys.exit(1)
+ path = _preset_path(args.name)
+ if not path.exists():
+ print(f"Error: Preset '{args.name}' not found.", file=sys.stderr)
+ sys.exit(1)
+ path.unlink()
+ print(f"Preset '{args.name}' deleted.")
+
+
+def main():
+ parser = argparse.ArgumentParser(description="Claude Banana Brand/Style Presets")
+ sub = parser.add_subparsers(dest="command", required=True)
+
+ # list
+ sub.add_parser("list", help="List available presets")
+
+ # show
+ p_show = sub.add_parser("show", help="Show preset details")
+ p_show.add_argument("name", help="Preset name")
+
+ # create
+ p_create = sub.add_parser("create", help="Create a new preset")
+ p_create.add_argument("name", help="Preset name (e.g., tech-saas, luxury-brand)")
+ p_create.add_argument("--colors", default="", help="Comma-separated hex colors")
+ p_create.add_argument("--style", default="", help="Visual style description")
+ p_create.add_argument("--typography", default="", help="Typography description")
+ p_create.add_argument("--lighting", default="", help="Lighting description")
+ p_create.add_argument("--mood", default="", help="Mood/emotion description")
+ p_create.add_argument("--description", default="", help="Brief preset description")
+ p_create.add_argument("--ratio", default="16:9", help="Default aspect ratio")
+ p_create.add_argument("--resolution", default="2K", help="Default resolution")
+
+ # delete
+ p_delete = sub.add_parser("delete", help="Delete a preset")
+ p_delete.add_argument("name", help="Preset name")
+ p_delete.add_argument("--confirm", action="store_true", help="Confirm deletion")
+
+ args = parser.parse_args()
+ cmds = {"list": cmd_list, "show": cmd_show, "create": cmd_create, "delete": cmd_delete}
+ cmds[args.command](args)
+
+
+if __name__ == "__main__":
+ main()
diff --git a/extensions/banana/scripts/setup_mcp.py b/extensions/banana/scripts/setup_mcp.py
new file mode 100755
index 0000000..f685a4a
--- /dev/null
+++ b/extensions/banana/scripts/setup_mcp.py
@@ -0,0 +1,151 @@
+#!/usr/bin/env python3
+"""
+Setup script for Claude Banana MCP server in Claude Code.
+
+Configures @ycse/nanobanana-mcp in Claude Code's settings.json
+with the user's Google AI API key.
+
+Usage:
+ python3 setup_mcp.py # Interactive (prompts for key)
+ python3 setup_mcp.py --key YOUR_KEY # Non-interactive
+ python3 setup_mcp.py --check # Verify existing setup
+ python3 setup_mcp.py --remove # Remove MCP config
+ python3 setup_mcp.py --help # Show usage
+"""
+
+import json
+import sys
+import os
+from pathlib import Path
+
+SETTINGS_PATH = Path.home() / ".claude" / "settings.json"
+MCP_NAME = "nanobanana-mcp"
+MCP_PACKAGE = "@ycse/nanobanana-mcp"
+DEFAULT_MODEL = "gemini-3.1-flash-image-preview"
+
+
+def load_settings() -> dict:
+ """Load Claude Code settings.json."""
+ if not SETTINGS_PATH.exists():
+ return {}
+ with open(SETTINGS_PATH, "r") as f:
+ return json.load(f)
+
+
+def save_settings(settings: dict) -> None:
+ """Save Claude Code settings.json."""
+ SETTINGS_PATH.parent.mkdir(parents=True, exist_ok=True)
+ with open(SETTINGS_PATH, "w") as f:
+ json.dump(settings, f, indent=2)
+ print(f"Settings saved to {SETTINGS_PATH}")
+
+
+def check_setup() -> bool:
+ """Check if MCP is already configured."""
+ settings = load_settings()
+ servers = settings.get("mcpServers", {})
+ if MCP_NAME in servers:
+ env = servers[MCP_NAME].get("env", {})
+ key = env.get("GOOGLE_AI_API_KEY", "")
+ masked = key[:8] + "..." + key[-4:] if len(key) > 12 else "(not set)"
+ print(f"MCP server '{MCP_NAME}' is configured.")
+ print(f" Package: {MCP_PACKAGE}")
+ print(f" API Key: {masked}")
+ print(f" Model: {env.get('NANOBANANA_MODEL', DEFAULT_MODEL)}")
+ return True
+ print(f"MCP server '{MCP_NAME}' is NOT configured.")
+ return False
+
+
+def remove_mcp() -> None:
+ """Remove MCP configuration."""
+ settings = load_settings()
+ servers = settings.get("mcpServers", {})
+ if MCP_NAME in servers:
+ del servers[MCP_NAME]
+ settings["mcpServers"] = servers
+ save_settings(settings)
+ print(f"Removed '{MCP_NAME}' from Claude Code settings.")
+ else:
+ print(f"'{MCP_NAME}' not found in settings.")
+
+
+def setup_mcp(api_key: str) -> None:
+ """Configure MCP server in Claude Code settings."""
+ if not api_key or not api_key.strip():
+ print("Error: API key cannot be empty.")
+ sys.exit(1)
+
+ api_key = api_key.strip()
+ settings = load_settings()
+
+ if "mcpServers" not in settings:
+ settings["mcpServers"] = {}
+
+ settings["mcpServers"][MCP_NAME] = {
+ "command": "npx",
+ "args": ["-y", MCP_PACKAGE],
+ "env": {
+ "GOOGLE_AI_API_KEY": api_key,
+ "NANOBANANA_MODEL": DEFAULT_MODEL,
+ },
+ }
+
+ save_settings(settings)
+ print(f"\nMCP server '{MCP_NAME}' configured successfully!")
+ print(f" Package: {MCP_PACKAGE}")
+ print(f" Model: {DEFAULT_MODEL}")
+ print(f"\nRestart Claude Code for changes to take effect.")
+ print(f"Generated images will be saved to: ~/Documents/nanobanana_generated/")
+
+
+def main() -> None:
+ args = sys.argv[1:]
+
+ if "--help" in args or "-h" in args:
+ print("Usage: python3 setup_mcp.py [OPTIONS]")
+ print()
+ print("Options:")
+ print(" --key KEY Provide API key non-interactively")
+ print(" --check Verify existing setup")
+ print(" --remove Remove MCP configuration")
+ print(" --help, -h Show this help message")
+ print()
+ print("Get a free API key at: https://aistudio.google.com/apikey")
+ sys.exit(0)
+
+ if "--check" in args:
+ check_setup()
+ return
+
+ if "--remove" in args:
+ remove_mcp()
+ return
+
+ # Get API key
+ api_key = None
+ for i, arg in enumerate(args):
+ if arg == "--key" and i + 1 < len(args):
+ api_key = args[i + 1]
+ break
+
+ if not api_key:
+ # Check environment
+ api_key = os.environ.get("GOOGLE_AI_API_KEY")
+
+ if not api_key:
+ print("Claude Banana — MCP Setup")
+ print("=" * 40)
+ print(f"\nGet your free API key at: https://aistudio.google.com/apikey")
+ print()
+ try:
+ api_key = input("Enter your Google AI API key: ")
+ except (EOFError, KeyboardInterrupt):
+ print("\nError: No input received. Provide a key with --key or set GOOGLE_AI_API_KEY env var.")
+ sys.exit(1)
+
+ setup_mcp(api_key)
+
+
+if __name__ == "__main__":
+ main()
diff --git a/extensions/banana/scripts/validate_setup.py b/extensions/banana/scripts/validate_setup.py
new file mode 100755
index 0000000..173e6cb
--- /dev/null
+++ b/extensions/banana/scripts/validate_setup.py
@@ -0,0 +1,133 @@
+#!/usr/bin/env python3
+"""
+Validate that the Claude Banana MCP server is properly configured.
+
+Checks:
+1. Claude Code settings.json has the MCP entry
+2. API key is present
+3. Node.js/npx is available
+4. Output directory exists or can be created
+
+Usage:
+ python3 validate_setup.py
+"""
+
+import json
+import shutil
+import sys
+from pathlib import Path
+
+SETTINGS_PATH = Path.home() / ".claude" / "settings.json"
+MCP_NAME = "nanobanana-mcp"
+OUTPUT_DIR = Path.home() / "Documents" / "nanobanana_generated"
+
+
+def check(label: str, passed: bool, detail: str = "") -> bool:
+ status = "PASS" if passed else "FAIL"
+ msg = f" [{status}] {label}"
+ if detail:
+ msg += f" — {detail}"
+ print(msg)
+ return passed
+
+
+def main() -> int:
+ print("Claude Banana — Setup Validation")
+ print("=" * 40)
+ results = []
+
+ # 1. Settings file exists
+ results.append(check(
+ "Claude Code settings.json exists",
+ SETTINGS_PATH.exists(),
+ str(SETTINGS_PATH),
+ ))
+
+ if not SETTINGS_PATH.exists():
+ print("\nCannot continue without settings.json.")
+ return 1
+
+ # 2. Load and parse settings
+ try:
+ with open(SETTINGS_PATH) as f:
+ settings = json.load(f)
+ results.append(check("settings.json is valid JSON", True))
+ except json.JSONDecodeError as e:
+ results.append(check("settings.json is valid JSON", False, str(e)))
+ return 1
+
+ # 3. MCP entry exists
+ servers = settings.get("mcpServers", {})
+ has_mcp = MCP_NAME in servers
+ results.append(check(f"MCP server '{MCP_NAME}' configured", has_mcp))
+
+ if has_mcp:
+ mcp = servers[MCP_NAME]
+
+ # 4. Command is npx
+ results.append(check(
+ "Command is 'npx'",
+ mcp.get("command") == "npx",
+ mcp.get("command", "(missing)"),
+ ))
+
+ # 5. Package is correct
+ args = mcp.get("args", [])
+ has_pkg = "@ycse/nanobanana-mcp" in args
+ results.append(check(
+ "Package is @ycse/nanobanana-mcp",
+ has_pkg,
+ str(args),
+ ))
+
+ # 6. API key present
+ env = mcp.get("env", {})
+ key = env.get("GOOGLE_AI_API_KEY", "")
+ results.append(check(
+ "GOOGLE_AI_API_KEY is set",
+ bool(key),
+ f"{key[:8]}...{key[-4:]}" if len(key) > 12 else "(empty or short)",
+ ))
+
+ # 7. Model configured
+ model = env.get("NANOBANANA_MODEL", "")
+ results.append(check(
+ "NANOBANANA_MODEL is set",
+ bool(model),
+ model or "(not set, will use package default)",
+ ))
+
+ # 8. Node.js/npx available
+ has_npx = shutil.which("npx") is not None
+ results.append(check(
+ "npx is available in PATH",
+ has_npx,
+ shutil.which("npx") or "not found",
+ ))
+
+ # 9. Output directory
+ if OUTPUT_DIR.exists():
+ results.append(check("Output directory exists", True, str(OUTPUT_DIR)))
+ else:
+ try:
+ OUTPUT_DIR.mkdir(parents=True, exist_ok=True)
+ results.append(check("Output directory created", True, str(OUTPUT_DIR)))
+ except OSError as e:
+ results.append(check("Output directory writable", False, str(e)))
+
+ # Summary
+ passed = sum(1 for r in results if r)
+ total = len(results)
+ print(f"\n{'=' * 40}")
+ print(f"Results: {passed}/{total} checks passed")
+
+ if passed == total:
+ print("Status: Ready to generate images!")
+ return 0
+ else:
+ print("Status: Some checks failed. Fix the issues above.")
+ return 1
+
+
+if __name__ == "__main__":
+ sys.exit(main())
diff --git a/extensions/banana/skills/seo-image-gen/SKILL.md b/extensions/banana/skills/seo-image-gen/SKILL.md
new file mode 100644
index 0000000..6e14118
--- /dev/null
+++ b/extensions/banana/skills/seo-image-gen/SKILL.md
@@ -0,0 +1,166 @@
+---
+name: seo-image-gen
+description: "AI image generation for SEO assets — OG/social preview images, blog hero images, schema images, product photography, infographics. Powered by Gemini via nanobanana-mcp. Requires banana extension installed. Use when user says \"generate image\", \"OG image\", \"social preview\", \"hero image\", \"blog image\", \"product photo\", \"infographic\", \"seo image\", \"create visual\", \"image-gen\", \"favicon\", \"schema image\", \"pinterest pin\", \"generate visual\", \"banner\", or \"thumbnail\"."
+argument-hint: "[og|hero|product|infographic|custom|batch] <description>"
+---
+
+# SEO Image Gen — AI Image Generation for SEO Assets (Extension)
+
+Generate production-ready images for SEO use cases using Gemini's image generation
+via the banana Creative Director pipeline. Maps SEO needs to optimized domain modes,
+aspect ratios, and resolution defaults.
+
+## Architecture Note
+
+This skill has two components with distinct roles:
+- **SKILL.md** (this file) — Handles interactive `/seo image-gen` commands for generating images
+- **Agent** (`agents/seo-image-gen.md`) — Audit-only analyst spawned during `/seo audit` to assess existing OG/social images and produce a generation plan (never auto-generates)
+
+## Prerequisites
+
+This skill requires the banana extension to be installed:
+```bash
+./extensions/banana/install.sh
+```
+
+**Check availability:** Before using any image generation tool, verify the MCP server
+is connected by checking if `gemini_generate_image` or `set_aspect_ratio` tools are
+available. If tools are not available, inform the user the extension is not installed
+and provide install instructions.
+
+## Quick Reference
+
+| Command | What it does |
+|---------|-------------|
+| `/seo image-gen og <description>` | Generate OG/social preview image (1200x630 feel) |
+| `/seo image-gen hero <description>` | Blog hero image (widescreen, dramatic) |
+| `/seo image-gen product <description>` | Product photography (clean, white BG) |
+| `/seo image-gen infographic <description>` | Infographic visual (vertical, data-heavy) |
+| `/seo image-gen custom <description>` | Custom image with full Creative Director pipeline |
+| `/seo image-gen batch <description> [N]` | Generate N variations (default: 3) |
+
+## SEO Image Use Cases
+
+Each use case maps to pre-configured banana parameters:
+
+| Use Case | Aspect Ratio | Resolution | Domain Mode | Notes |
+|----------|-------------|------------|-------------|-------|
+| **OG/Social Preview** | `16:9` | `1K` | Product or UI/Web | Clean, professional, text-friendly |
+| **Blog Hero** | `16:9` | `2K` | Cinema or Editorial | Dramatic, atmospheric, editorial quality |
+| **Schema Image** | `4:3` | `1K` | Product | Clean, descriptive, schema ImageObject |
+| **Social Square** | `1:1` | `1K` | UI/Web | Platform-optimized square |
+| **Product Photo** | `4:3` | `2K` | Product | White background, studio lighting |
+| **Infographic** | `2:3` | `4K` | Infographic | Data-heavy, vertical layout |
+| **Favicon/Icon** | `1:1` | `512` | Logo | Minimal, scalable, recognizable |
+| **Pinterest Pin** | `2:3` | `2K` | Editorial | Tall vertical card |
+
+## Generation Pipeline
+
+For every generation request:
+
+1. **Identify use case** from command or context (og, hero, product, etc.)
+2. **Apply SEO defaults** from the use cases table above
+3. **Set aspect ratio** via `set_aspect_ratio` MCP tool
+4. **Construct Reasoning Brief** using the banana Creative Director pipeline:
+ - Load `references/prompt-engineering.md` for the 6-component system
+ - Apply domain mode emphasis (Subject 30%, Style 25%, Context 15%, etc.)
+ - Be SPECIFIC and VISCERAL — describe what the camera sees
+5. **Generate** via `gemini_generate_image` MCP tool
+6. **Post-generation SEO checklist** (see below)
+
+### Check for Presets
+
+If the user mentions a brand or has SEO presets configured:
+```bash
+python3 ~/.claude/skills/seo-image-gen/scripts/presets.py list
+```
+Load matching preset and apply as defaults. Also check `references/seo-image-presets.md`
+for SEO-specific preset templates.
+
+## Post-Generation SEO Checklist
+
+After every successful generation, guide the user on:
+
+1. **Alt text** — Write descriptive, keyword-rich alt text for the generated image
+2. **File naming** — Rename to SEO-friendly format: `keyword-description-widthxheight.webp`
+3. **WebP conversion** — Convert to WebP for optimal page speed:
+ ```bash
+ magick output.png -quality 85 output.webp
+ ```
+4. **File size** — Target under 200KB for hero images, under 100KB for thumbnails
+5. **Schema markup** — Suggest `ImageObject` schema for the generated image:
+ ```json
+ {
+ "@type": "ImageObject",
+ "url": "https://example.com/images/keyword-description.webp",
+ "width": 1200,
+ "height": 630,
+ "caption": "Descriptive caption with target keyword"
+ }
+ ```
+6. **OG meta tags** — For social preview images, remind about:
+ ```html
+ <meta property="og:image" content="https://example.com/images/og-image.webp" />
+ <meta property="og:image:width" content="1200" />
+ <meta property="og:image:height" content="630" />
+ <meta property="og:image:alt" content="Descriptive alt text" />
+ ```
+
+## Cost Awareness
+
+Image generation costs money. Be transparent:
+- Show estimated cost before generating (especially for batch)
+- Log every generation: `python3 ~/.claude/skills/seo-image-gen/scripts/cost_tracker.py log --model MODEL --resolution RES --prompt "brief"`
+- Run `cost_tracker.py summary` if user asks about usage
+
+Approximate costs (gemini-3.1-flash):
+- 512: ~$0.02/image
+- 1K resolution: ~$0.04/image
+- 2K resolution: ~$0.08/image
+- 4K resolution: ~$0.16/image
+
+## Model Routing
+
+| Scenario | Model | Why |
+|----------|-------|-----|
+| OG images, social previews | `gemini-3.1-flash-image-preview` @ 1K | Fast, cost-effective |
+| Hero images, product photos | `gemini-3.1-flash-image-preview` @ 2K | Quality + detail |
+| Infographics with text | `gemini-3.1-flash-image-preview` @ 2K, thinking: high | Better text rendering |
+| Quick drafts | `gemini-2.5-flash-image` @ 512 | Rapid iteration |
+
+## Error Handling
+
+| Error | Resolution |
+|-------|-----------|
+| MCP not configured | Run `./extensions/banana/install.sh` |
+| API key invalid | New key at https://aistudio.google.com/apikey |
+| Rate limited (429) | Wait 60s, retry. Free tier: ~10 RPM / ~500 RPD |
+| `IMAGE_SAFETY` | Rephrase prompt — see `references/prompt-engineering.md` Safety section |
+| MCP unavailable | Fall back: `python3 ~/.claude/skills/seo-image-gen/scripts/generate.py --prompt "..." --aspect-ratio "16:9"` |
+| Extension not installed | Show install instructions: `./extensions/banana/install.sh` |
+
+## Cross-Skill Integration
+
+- **seo-images** (analysis) feeds into **seo-image-gen** (generation): audit results from `/seo images` identify missing or low-quality images; use those findings to drive `/seo image-gen` commands
+- **seo-audit** spawns the seo-image-gen **agent** (not this skill) to analyze OG/social images across the site and produce a prioritized generation plan
+- **seo-schema** can consume generated images: after generation, suggest `ImageObject` schema markup pointing to the new assets
+
+## Reference Documentation
+
+Load on-demand — do NOT load all at startup:
+- `references/prompt-engineering.md` — 6-component system, domain modes, templates
+- `references/gemini-models.md` — Model specs, rate limits, capabilities
+- `references/mcp-tools.md` — MCP tool parameters and responses
+- `references/post-processing.md` — ImageMagick/FFmpeg pipeline recipes
+- `references/cost-tracking.md` — Pricing, usage tracking
+- `references/presets.md` — Brand preset management
+- `references/seo-image-presets.md` — SEO-specific preset templates
+
+## Response Format
+
+After generating, always provide:
+1. **Image path** — where it was saved
+2. **Crafted prompt** — show what was sent to the API (educational)
+3. **Settings** — model, aspect ratio, resolution
+4. **SEO checklist** — alt text suggestion, file naming, WebP conversion
+5. **Schema snippet** — ImageObject or og:image markup if applicable
diff --git a/extensions/banana/uninstall.sh b/extensions/banana/uninstall.sh
new file mode 100755
index 0000000..d036bde
--- /dev/null
+++ b/extensions/banana/uninstall.sh
@@ -0,0 +1,43 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+main() {
+ echo "→ Uninstalling Banana Image Generation extension..."
+
+ # Remove skill (includes copied scripts and references)
+ rm -rf "${HOME}/.claude/skills/seo-image-gen"
+
+ # Remove agent
+ rm -f "${HOME}/.claude/agents/seo-image-gen.md"
+
+ # Ask before removing MCP server (user may use standalone banana skill)
+ SETTINGS_FILE="${HOME}/.claude/settings.json"
+ if [ -f "${SETTINGS_FILE}" ]; then
+ # Check if standalone banana skill still exists
+ if [ -d "${HOME}/.claude/skills/banana" ]; then
+ echo " ℹ Standalone banana skill detected at ~/.claude/skills/banana/"
+ echo " ℹ Keeping nanobanana-mcp in settings.json (used by standalone skill)"
+ else
+ # No standalone skill, safe to remove MCP
+ python3 -c "
+import json, os
+settings_path = '${SETTINGS_FILE}'
+with open(settings_path, 'r') as f:
+ settings = json.load(f)
+if 'mcpServers' in settings and 'nanobanana-mcp' in settings['mcpServers']:
+ del settings['mcpServers']['nanobanana-mcp']
+ if not settings['mcpServers']:
+ del settings['mcpServers']
+ with open(settings_path, 'w') as f:
+ json.dump(settings, f, indent=2)
+ print(' ✓ Removed nanobanana-mcp from settings.json')
+else:
+ print(' ✓ No nanobanana-mcp entry in settings.json')
+" 2>/dev/null || echo " ⚠ Could not auto-remove MCP config. Remove 'nanobanana-mcp' from ~/.claude/settings.json manually."
+ fi
+ fi
+
+ echo "✓ Banana Image Generation extension uninstalled."
+}
+
+main "$@"
diff --git a/seo/SKILL.md b/seo/SKILL.md
index 76db4c5..c5aaf3c 100644
--- a/seo/SKILL.md
+++ b/seo/SKILL.md
@@ -37,6 +37,7 @@ e-commerce, publishers, agencies). Orchestrates 12 specialized sub-skills and 7
| `/seo competitor-pages [url\|generate]` | Competitor comparison page generation |
| `/seo hreflang [url]` | Hreflang/i18n SEO audit and generation |
| `/seo dataforseo [command]` | Live SEO data via DataForSEO (extension) |
+| `/seo image-gen [use-case] <description>` | AI image generation for SEO assets (extension) |
## Orchestration Logic
@@ -98,7 +99,7 @@ Weighted aggregate of all categories:
## Sub-Skills
-This skill orchestrates 12 specialized sub-skills (+ 1 extension):
+This skill orchestrates 12 specialized sub-skills (+ 2 extensions):
1. **seo-audit** — Full website audit with parallel delegation
2. **seo-page** — Deep single-page analysis
@@ -113,6 +114,7 @@ This skill orchestrates 12 specialized sub-skills (+ 1 extension):
11. **seo-competitor-pages** — Competitor comparison page generation
12. **seo-hreflang** — Hreflang/i18n SEO audit and generation
13. **seo-dataforseo** — Live SEO data via DataForSEO MCP (extension)
+14. **seo-image-gen** — AI image generation for SEO assets via Gemini (extension)
## Subagents
@@ -125,3 +127,4 @@ For parallel analysis during audits:
- `seo-visual` — Screenshots, mobile testing, above-fold
- `seo-geo` — AI crawler access, llms.txt, citability, brand mention signals
- `seo-dataforseo` — Live SERP, keyword, backlink, local SEO data (extension, optional)
+- `seo-image-gen` — SEO image audit and generation plan (extension, optional)