Nano Banana Pro Prompts Recommend Skill

https://github.com/youmind-openlab/nano-banana-pro-prompts-recommend-skill
Admin
A skill for Claude Code that intelligently recommends from 6000+ curated image generation prompts,...
Tokens:5,542
Snippets:47
Trust Score:6.6
Update:8 months ago
Show doc for...
Tokens:
Raw
### Install Skill via CLI

Source: https://context7.com/youmind-openlab/nano-banana-pro-prompts-recommend-skill/llms.txt

Installs the Nano Banana Pro Prompts Recommend Skill using the skills command-line interface. This is the recommended installation method.

```bash
npx skills i YouMind-OpenLab/nano-banana-pro-prompts-recommend-skill
```

--------------------------------

### Install Skill via Claude Code Plugin Marketplace

Source: https://context7.com/youmind-openlab/nano-banana-pro-prompts-recommend-skill/llms.txt

Installs the Nano Banana Pro Prompts Recommend Skill directly from the Claude Code Plugin Marketplace using a specific command.

```bash
/plugin marketplace add YouMind-OpenLab/nano-banana-pro-prompts-recommend-skill
```

--------------------------------

### Install and Update Skill via OpenSkills CLI

Source: https://context7.com/youmind-openlab/nano-banana-pro-prompts-recommend-skill/llms.txt

Provides commands to install and update the Nano Banana Pro Prompts Recommend Skill using the openskills command-line tool. Ensures the skill is up-to-date.

```bash
# Install the skill
npx openskills install YouMind-OpenLab/nano-banana-pro-prompts-recommend-skill

# Update to latest version
npx openskills update nano-banana-pro-prompts-recommend-skill
```

--------------------------------

### Search Prompts with Grep Tool

Source: https://context7.com/youmind-openlab/nano-banana-pro-prompts-recommend-skill/llms.txt

Demonstrates how to use the Grep tool to search for specific patterns within prompt reference files, optimizing token usage by avoiding loading entire files. Includes examples for single and multiple category searches, and fallback searches.

```bash
# Search for cartoon-style avatar prompts
Grep pattern="cartoon" path="references/profile-avatar.json"

# Search for product photography prompts
Grep pattern="white background" path="references/ecommerce-main-image.json"

# Search across multiple categories
Grep pattern="minimalist" path="references/app-web-design.json"
Grep pattern="minimalist" path="references/poster-flyer.json"

# Fallback to others.json if no match in primary category
Grep pattern="landscape" path="references/others.json"
```

--------------------------------

### Generate Reference Data from CMS (Bash)

Source: https://context7.com/youmind-openlab/nano-banana-pro-prompts-recommend-skill/llms.txt

Bash commands to set up environment variables for connecting to the PayloadCMS and install necessary dependencies using pnpm for generating reference data.

```bash
# Setup environment variables
 echo "CMS_HOST=your_cms_host" >> .env
 echo "CMS_API_KEY=your_api_key" >> .env

# Install dependencies
 pnpm install

```

--------------------------------

### GitHub Actions Workflow for Automated Updates (YAML)

Source: https://context7.com/youmind-openlab/nano-banana-pro-prompts-recommend-skill/llms.txt

This YAML workflow automates the process of checking out the repository, setting up Node.js and pnpm, installing dependencies, running the generation script, and committing/pushing any changes to the references. It is scheduled to run twice daily.

```yaml
# .github/workflows/generate-references.yml
name: Generate References

on:
  schedule:
    - cron: '0 0 * * *'   # Run at 00:00 UTC
    - cron: '0 12 * * *'  # Run at 12:00 UTC
  workflow_dispatch:       # Manual trigger

jobs:
  generate:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-node@v4
        with:
          node-version: '20'
      - uses: pnpm/action-setup@v4
      - run: pnpm install --frozen-lockfile
      - run: pnpm run generate
        env:
          CMS_HOST: ${{ secrets.CMS_HOST }}
          CMS_API_KEY: ${{ secrets.CMS_API_KEY }}
      - name: Commit and push changes
        run: |
          git config --local user.email "github-actions[bot]@users.noreply.github.com"
          git config --local user.name "github-actions[bot]"
          git add references/ SKILL.md
          git commit -m "chore: update references $(date -u '+%Y-%m-%d %H:%M UTC')"
          git push

```

--------------------------------

### Prompt Data Structure (JSON)

Source: https://context7.com/youmind-openlab/nano-banana-pro-prompts-recommend-skill/llms.txt

Defines the JSON structure for each prompt within the reference files. Includes content, title, description, source media, and a flag for reference image needs.

```json
{
  "content": "Conceptual storytelling visual for a modern brand narrative.\nAbstract progression elements,\nsense of movement and flow,\nsoft cinematic lighting,\noptimistic, forward-moving tone,\ncrafted to communicate journey and growth.",
  "title": "Conceptual brand storytelling visual prompt",
  "description": "A concise prompt for generating a conceptual visual that communicates a brand narrative focused on journey and growth.",
  "sourceMedia": [
    "https://cms-assets.youmind.com/media/example_image.jpg"
  ],
  "needReferenceImages": false
}
```

--------------------------------

### Reference File Category Mapping (TypeScript)

Source: https://context7.com/youmind-openlab/nano-banana-pro-prompts-recommend-skill/llms.txt

A TypeScript object mapping JSON filenames in the 'references/' directory to human-readable category names and prompt counts. Used for organizing and searching prompts.

```typescript
// Category mapping for prompt search
const categoryFiles = {
  "profile-avatar.json": "Profile / Avatar (744 prompts)",
  "social-media-post.json": "Social Media Post (3973 prompts)",
  "product-marketing.json": "Product Marketing (2047 prompts)",
  "infographic-edu-visual.json": "Infographic / Edu Visual (359 prompts)",
  "poster-flyer.json": "Poster / Flyer (345 prompts)",
  "game-asset.json": "Game Asset (217 prompts)",
  "comic-storyboard.json": "Comic / Storyboard (215 prompts)",
  "ecommerce-main-image.json": "E-commerce Main Image (215 prompts)",
  "youtube-thumbnail.json": "YouTube Thumbnail (117 prompts)",
  "app-web-design.json": "App / Web Design (116 prompts)",
  "others.json": "Uncategorized (677 prompts)"
};
```

--------------------------------

### Generate Reference Files Script (TypeScript)

Source: https://context7.com/youmind-openlab/nano-banana-pro-prompts-recommend-skill/llms.txt

The core TypeScript script responsible for fetching prompts from a CMS API, transforming them into the desired output format, and saving them as reference files. It handles API requests, data transformation, and file system operations.

```typescript
// scripts/generate-references.ts - Core generation logic
import "dotenv/config";
import { stringify } from "qs-esm";
import * as fs from "fs";
import * as path from "path";

interface OutputPrompt {
  content: string;
  title: string;
  description: string;
  sourceMedia: string[];
  needReferenceImages: boolean;
}

// Fetch all prompts from CMS API
async function fetchAllPrompts(): Promise<CMSPrompt[]> {
  const query = {
    limit: 100,
    page: 1,
    sort: ["sort", "-sourcePublishedAt"].join(","),
    depth: 2,
    locale: "en",
    where: {
      model: { equals: "nano-banana-pro" }
    }
  };

  const response = await fetch(`${CMS_HOST}/api/prompts${stringify(query, { addQueryPrefix: true })}`, {
    headers: {
      Authorization: `users API-Key ${CMS_API_KEY}`,
      "Content-Type": "application/json"
    }
  });

  return (await response.json()).docs;
}

// Transform CMS prompt to output format
function transformToOutputPrompt(item: CMSPrompt): OutputPrompt | null {
  const sourceMedia = item.media?.map(m => m.url).filter(Boolean) || item.sourceMedia || [];
  if (sourceMedia.length === 0) return null;

  return {
    content: item.translatedContent || item.content,
    title: item.title,
    description: item.description,
    sourceMedia,
    needReferenceImages: item.needReferenceImages ?? false
  };
}

```

--------------------------------

### Category Signal Mapping (TypeScript)

Source: https://context7.com/youmind-openlab/nano-banana-pro-prompts-recommend-skill/llms.txt

A TypeScript object mapping user request keywords (signals) to their corresponding prompt category JSON filenames. Facilitates routing user requests to the correct prompt data.

```typescript
const signalMapping = {
  // User request signals -> Target file
  "avatar, profile picture, headshot, portrait, selfie": "profile-avatar.json",
  "post, instagram, twitter, facebook, social, viral": "social-media-post.json",
  "infographic, diagram, educational, data visualization, chart": "infographic-edu-visual.json",
  "thumbnail, youtube, video cover, click-bait": "youtube-thumbnail.json",
  "comic, manga, storyboard, panel, cartoon story": "comic-storyboard.json",
  "product, marketing, advertisement, promo, campaign": "product-marketing.json",
  "e-commerce, product photo, white background, listing": "ecommerce-main-image.json",
  "game, asset, sprite, character design, item": "game-asset.json",
  "poster, flyer, banner, announcement, event": "poster-flyer.json",
  "app, UI, website, interface, mockup": "app-web-design.json"
};
```

--------------------------------

### Prompt Data Structure JSON

Source: https://github.com/youmind-openlab/nano-banana-pro-prompts-recommend-skill/blob/main/SKILL.md

Defines the structure for prompt data used in image generation. It includes the main prompt content, a title, a description of the generated output, optional source media URLs, and a flag indicating if reference images are needed.

```json
{
  "content": "English prompt text for image generation",
  "title": "Prompt title",
  "description": "What this prompt creates",
  "sourceMedia": ["image_url_1", "image_url_2"],
  "needReferenceImages": false
}
```
Tokens:
Raw