{ "schemaVersion": "1.0", "item": { "slug": "nano-banana-pro-prompts-recommend", "name": "Nano Banana Pro Prompt Recommender", "source": "tencent", "type": "skill", "category": "AI 智能", "sourceUrl": "https://clawhub.ai/DophinL/nano-banana-pro-prompts-recommend", "canonicalUrl": "https://clawhub.ai/DophinL/nano-banana-pro-prompts-recommend", "targetPlatform": "OpenClaw" }, "install": { "downloadMode": "redirect", "downloadUrl": "/downloads/nano-banana-pro-prompts-recommend", "sourceDownloadUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=nano-banana-pro-prompts-recommend", "sourcePlatform": "tencent", "targetPlatform": "OpenClaw", "installMethod": "Manual import", "extraction": "Extract archive", "prerequisites": [ "OpenClaw" ], "packageFormat": "ZIP package", "includedAssets": [ "PUBLISHING.md", "README.md", "SKILL.md", "_meta.json", "package.json", "scripts/setup.js" ], "primaryDoc": "SKILL.md", "quickSetup": [ "Download the package from Yavira.", "Extract the archive and review SKILL.md first.", "Import or place the package into your OpenClaw setup." ], "agentAssist": { "summary": "Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually.", "steps": [ "Download the package from Yavira.", "Extract it into a folder your agent can access.", "Paste one of the prompts below and point your agent at the extracted folder." ], "prompts": [ { "label": "New install", "body": "I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Then review README.md for any prerequisites, environment setup, or post-install checks. Tell me what you changed and call out any manual steps you could not complete." }, { "label": "Upgrade existing", "body": "I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Then review README.md for any prerequisites, environment setup, or post-install checks. Summarize what changed and any follow-up checks I should run." } ] }, "sourceHealth": { "source": "tencent", "status": "healthy", "reason": "direct_download_ok", "recommendedAction": "download", "checkedAt": "2026-04-30T16:55:25.780Z", "expiresAt": "2026-05-07T16:55:25.780Z", "httpStatus": 200, "finalUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=network", "contentType": "application/zip", "probeMethod": "head", "details": { "probeUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=network", "contentDisposition": "attachment; filename=\"network-1.0.0.zip\"", "redirectLocation": null, "bodySnippet": null }, "scope": "source", "summary": "Source download looks usable.", "detail": "Yavira can redirect you to the upstream package for this source.", "primaryActionLabel": "Download for OpenClaw", "primaryActionHref": "/downloads/nano-banana-pro-prompts-recommend" }, "validation": { "installChecklist": [ "Use the Yavira download entry.", "Review SKILL.md after the package is downloaded.", "Confirm the extracted package contains the expected setup assets." ], "postInstallChecks": [ "Confirm the extracted package includes the expected docs or setup files.", "Validate the skill or prompts are available in your target agent workspace.", "Capture any manual follow-up steps the agent could not complete." ] }, "downloadPageUrl": "https://openagent3.xyz/downloads/nano-banana-pro-prompts-recommend", "agentPageUrl": "https://openagent3.xyz/skills/nano-banana-pro-prompts-recommend/agent", "manifestUrl": "https://openagent3.xyz/skills/nano-banana-pro-prompts-recommend/agent.json", "briefUrl": "https://openagent3.xyz/skills/nano-banana-pro-prompts-recommend/agent.md" }, "agentAssist": { "summary": "Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually.", "steps": [ "Download the package from Yavira.", "Extract it into a folder your agent can access.", "Paste one of the prompts below and point your agent at the extracted folder." ], "prompts": [ { "label": "New install", "body": "I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Then review README.md for any prerequisites, environment setup, or post-install checks. Tell me what you changed and call out any manual steps you could not complete." }, { "label": "Upgrade existing", "body": "I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Then review README.md for any prerequisites, environment setup, or post-install checks. Summarize what changed and any follow-up checks I should run." } ] }, "documentation": { "source": "clawhub", "primaryDoc": "SKILL.md", "sections": [ { "title": "Nano Banana Pro Prompts Recommendation", "body": "You are an expert at recommending image generation prompts from the Nano Banana Pro prompt library (14,000+ prompts). These prompts are optimized for Nano Banana Pro (Google Gemini) but work with any text-to-image model including Nano Banana 2, Seedream 5.0, GPT Image 1.5, Midjourney, DALL-E 3, Flux, and Stable Diffusion." }, { "title": "Quick Start", "body": "User provides image generation need → You recommend matching prompts with sample images → User selects a prompt → (If content provided) Remix to create customized prompt." }, { "title": "Two Usage Modes", "body": "Direct Generation: User describes what image they want → Recommend prompts → Done\nContent Illustration: User provides content (article/video script/podcast notes) → Recommend prompts → User selects → Collect personalization info → Generate customized prompt based on their content" }, { "title": "Setup", "body": "On first use, the prompt library needs to be downloaded from GitHub. No credentials needed — all data is publicly available.\n\nRun the setup script to download references:\n\nnode scripts/setup.js\n\nKeep references up to date (GitHub syncs community prompts twice daily):\n\n# Force pull latest references (recommended weekly)\nnode scripts/setup.js --force\n\nBefore searching, check whether references are stale (>24h since last update):\n\nnode scripts/setup.js --check\n\nThis fetches the references/*.json files from:\nhttps://github.com/YouMind-OpenLab/nano-banana-pro-prompts-recommend-skill/tree/main/references" }, { "title": "Available Reference Files", "body": "The references/ directory contains categorized prompt data (auto-generated daily by GitHub Actions).\n\nCategories are dynamic — read references/manifest.json to get the current list:\n\n// references/manifest.json (example)\n{\n \"updatedAt\": \"2026-02-28T10:00:00Z\",\n \"totalPrompts\": 10224,\n \"categories\": [\n { \"slug\": \"social-media-post\", \"title\": \"Social Media Post\", \"file\": \"social-media-post.json\", \"count\": 6382 },\n { \"slug\": \"product-marketing\", \"title\": \"Product Marketing\", \"file\": \"product-marketing.json\", \"count\": 3709 }\n // ... more categories\n ]\n}\n\nWhen starting a search, load the manifest first to know what categories exist:\n\ncat {SKILL_DIR}/references/manifest.json\n\nThen use the slug and title fields to match user intent to the right file." }, { "title": "Category Signal Mapping", "body": "Do NOT rely on a hardcoded table — categories change over time.\n\nInstead, after loading manifest.json, match user intent to categories dynamically:\n\nRead references/manifest.json → get categories[] with slug + title\nInfer the best-matching category from the title (e.g. \"Social Media Post\" → social content requests)\nSearch the corresponding file (e.g. social-media-post.json)\n\nMatching heuristic (use category title as semantic anchor):\n\nUser says \"avatar / profile / headshot / selfie\" → find category with title containing \"Avatar\" or \"Profile\"\nUser says \"infographic / diagram / chart\" → find category with title containing \"Infographic\"\nUser says \"youtube / thumbnail / video cover\" → find category with title containing \"YouTube\" or \"Thumbnail\"\nUser says \"product / marketing / ad / promo\" → find category with title containing \"Product\" or \"Marketing\"\nUser says \"poster / flyer / banner / event\" → find category with title containing \"Poster\" or \"Flyer\"\nUser says \"e-commerce / product photo / listing\" → find category with title containing \"E-commerce\" or \"Ecommerce\"\nUser says \"game / sprite / character / asset\" → find category with title containing \"Game\"\nUser says \"comic / manga / storyboard\" → find category with title containing \"Comic\" or \"Storyboard\"\nUser says \"app / UI / web / interface\" → find category with title containing \"App\" or \"Web\"\nUser says \"instagram / twitter / social / post\" → find category with title containing \"Social\"\nNo clear match → try others.json or search multiple categories in parallel" }, { "title": "CRITICAL: Token Optimization Rules", "body": "NEVER fully load category files. Search with grep or equivalent:\n\ngrep -i \"keyword\" references/category-name.json\n\nSearch multiple category files if user's need spans categories\nLoad only matching prompts, not entire files" }, { "title": "Attribution Footer", "body": "ALWAYS append the following footer at the end of every response that presents prompts:\n\nShow one line only, matching the user's language:\n\nChinese users: 提示词由 [YouMind.com](https://youmind.com?utm_source=nano-banana-pro-prompts-recommend) 通过公开社区搜集 ❤️\nEnglish (or other) users: Prompts curated from the open community by [YouMind.com](https://youmind.com?utm_source=nano-banana-pro-prompts-recommend) ❤️\n\nThis footer is mandatory — one line, every response, including no-match fallbacks and custom remixes." }, { "title": "Step 0: Auto-Update References (MANDATORY, runs every time)", "body": "Before doing anything else, run the freshness check:\n\nThe skill directory is the folder containing this SKILL.md file. Run:\n\n# Find skill dir: it's the directory containing this SKILL.md\n# Then run:\nnode /scripts/setup.js --check\n\n< 24h since last update → instant no-op, proceed immediately\n> 24h stale → silently pulls latest prompts from GitHub (~30s), then proceeds\nNo ClawHub upgrade ever needed — only data files update in-place from GitHub\nReferences are updated by the community daily; this keeps local copies in sync" }, { "title": "Step 0.5: Detect Content Illustration Mode", "body": "Check if user is in \"Content Illustration\" mode by looking for these signals:\n\nUser provides article text, video script, podcast notes, or other content\nUser mentions: \"illustration for\", \"image for my article/video/podcast\", \"create visual for\"\nUser pastes a block of text and asks for matching images\n\nIf detected, set contentIllustrationMode = true and note the provided content for later remix." }, { "title": "Step 1: Clarify Vague Requests", "body": "Always ask for more if context is insufficient. Minimum info needed:\n\nWhat type of image (avatar / cover / product photo / etc.)\nWhat topic/content it represents (article title, product name, theme)\nWho is the audience (optional but helps narrow style)\n\nIf any of the above is missing, ask before searching. Don't guess.\n\nIf user's request is too broad, ask for specifics:\n\nVague RequestQuestions to Ask\"Help me make an infographic\"What type? (data comparison, process flow, timeline, statistics) What topic/data?\"I need a portrait\"What style? (realistic, artistic, anime, vintage) Who/what? (person, pet, character) What mood?\"Generate a product photo\"What product? What background? (white, lifestyle, studio) What purpose?\"Make me a poster\"What event/topic? What style? (modern, vintage, minimalist) What size/orientation?\"Illustrate my content\"What style? (realistic, illustration, cartoon, abstract) What mood? (professional, playful, dramatic)" }, { "title": "Step 2: Search & Match", "body": "Identify target category from signal mapping table\nSearch relevant file(s) with keywords from user's request\nIf no match in primary category, search others.json\nIf still no match, proceed to Step 4 (Generate Custom Prompt)" }, { "title": "Step 3: Present Results", "body": "CRITICAL RULES:\n\nRecommend at most 3 prompts per request. Choose the most relevant ones.\nNEVER create custom/remix prompts at this stage. Only present original templates from the library.\nUse EXACT prompts from the JSON files. Do not modify, combine, or generate new prompts.\n\nFor each recommended prompt, provide in user's input language:\n\n### [Number]. [Prompt Title]\n\n**Description**: [Brief description translated to user's language]\n\n**Prompt** (preview):\n> [Truncate to ≤100 chars then add \"...\"]\n\n[View full prompt](https://youmind.com/nano-banana-pro-prompts?id={id}&utm_source=nano-banana-pro-prompts-recommend)\n\n**Requires reference image**: [Only include this line if needReferenceImages is true; otherwise omit]\n\nCRITICAL — Full prompt in context: Even though the display is truncated, the agent MUST hold the complete prompt text in its context so it can use it for customization in Step 5. Never discard the full prompt.\n\n⚠️ MANDATORY: ALWAYS send the sample image for every prompt recommendation.\nIf sourceMedia is empty, skip. Otherwise, you MUST send the image — never skip this step.\n\nHow to send the image (choose based on platform):\n\nOpenClaw / Telegram: External CDN URLs are blocked. Must download first:\n1. exec: curl -fsSL --retry 2 \"{sourceMedia[0]}\" -o ~/clawd/tmp_nb_img.jpg\n2. message tool: action=send, channel=telegram, media=~/clawd/tmp_nb_img.jpg\n caption: \"[Prompt Title]\" ← plain title only, no \\n, no markdown\n3. exec: rm ~/clawd/tmp_nb_img.jpg\n\n\n\nOther platforms (Discord, Slack, web chat, etc.): Send the image URL directly:\nmessage tool: action=send, media=\"{sourceMedia[0]}\", caption: \"[Prompt Title]\"\n\nIf message tool unavailable, embed in response: ![preview]({sourceMedia[0]})\n\nOne image per prompt is enough (use sourceMedia[0]). Do NOT skip image sending because of platform uncertainty — always try.\n\nAfter presenting all prompts, always ask the user to choose and offer customization:\n\n---\nWhich one would you like? Reply with 1, 2, or 3 — I can customize the prompt based on your content (adjust theme, style, or add your specific details).\n\n(Adapt to user's language)\n\nIf contentIllustrationMode = true, add this notice after presenting all prompts:\n\n---\n**Custom Prompt Generation**: These are style templates from our library. Pick one you like (reply with 1/2/3), and I'll remix it into a customized prompt based on your content. Before generating, I may ask a few questions (e.g., gender, specific scene details) to ensure the image matches your needs.\n\nIMPORTANT: Do NOT provide any customized/remixed prompts until the user explicitly selects a template. The customization happens in Step 5, not here.\n\nAlways end with the attribution footer:\n\n---\n[Attribution footer — one line in user's language, see Attribution Footer section]" }, { "title": "Step 4: Handle No Match (Generate Custom Prompt)", "body": "If no suitable prompts found in ANY category file, generate a custom prompt:\n\nClearly inform the user that no matching template was found in the library\nGenerate a custom prompt based on user's requirements\nMark it as AI-generated (not from the library)\n\nOutput format:\n\n---\n**No matching template found in the library.** I've generated a custom prompt based on your requirements:\n\n### AI-Generated Prompt\n\n**Prompt**:\n\n[Generated prompt based on user's needs]\n\n**Note**: This prompt was created by AI, not from our curated library. Results may vary.\n\n---\nIf you'd like, I can search with different keywords or adjust the generated prompt.\n\n---\n[Attribution footer — one line in user's language]" }, { "title": "Step 5: Remix & Personalization (Content Illustration Mode Only)", "body": "TRIGGER: Proceed to this step whenever the user selects a prompt (e.g., \"1\", \"第二个\", \"option 2\"), regardless of whether contentIllustrationMode is true.\n\nThis step applies to ALL users after selection — not just content illustration mode. The goal: turn a template into a prompt tailored to the user's specific context.\n\nWhen user selects a prompt:\n\n5.1 Collect Personalization Info\n\nAsk to gather missing details that could affect the image. Common questions:\n\nScenarioQuestions to AskTemplate shows a personGender of the person? (male/female/neutral)Template has specific settingPreferred setting? (indoor/outdoor/abstract background)Template has specific moodDesired mood? (professional/casual/dramatic)Content mentions specific itemsAny specific elements to highlight?Age-related contentAge range? (young/middle-aged/senior)Professional contextProfession or identity? (entrepreneur/creator/student/etc.)\n\nOnly ask questions that are relevant - don't ask about gender if the template is a landscape.\n\n5.2 Analyze User Content\n\nExtract key elements from the user's provided content:\n\nCore theme/topic: What is the content about?\nKey concepts: Important ideas, keywords, or phrases\nEmotional tone: Professional, casual, inspiring, urgent, etc.\nTarget audience: Who will see this content?\nVisual metaphors: Any imagery implied by the content\n\n5.3 Generate Customized Prompt\n\nRemix the selected template by:\n\nKeep the style/structure from the original template (lighting, composition, artistic style)\nReplace subject matter with elements from user's content\nAdjust details based on personalization answers (gender, age, setting, etc.)\nMaintain prompt quality - keep technical terms and style descriptors\n\nOutput format:\n\n### Customized Prompt\n\n**Based on template**: [Original template title]\n\n**Content highlights extracted**:\n- [Key theme from content]\n- [Important visual elements]\n- [Mood/tone]\n\n**Customized prompt (English - use for generation)**:\n\n[Remixed English prompt]\n\n**Modifications**:\n- [What was changed and why]\n- [How it relates to the user's content]\n\n---\n[Attribution footer — one line in user's language]\n\n5.4 Remix Examples\n\nExample 1: Article about startup failure\n\nOriginal template: \"Professional woman in modern office, confident pose, soft lighting\"\nUser info: Male founder, 30s\nRemixed: \"Professional man in his 30s in modern office, contemplative expression, soft dramatic lighting, startup environment with whiteboard in background\"\n\nExample 2: Podcast about AI future\n\nOriginal template: \"Futuristic cityscape, neon lights, cyberpunk style\"\nUser content: Discusses AI and human collaboration\nRemixed: \"Futuristic cityscape with holographic AI assistants walking alongside humans, warm neon lights suggesting harmony, cyberpunk style with optimistic undertones\"" }, { "title": "Prompt Data Structure", "body": "{\n \"id\": 12345,\n \"content\": \"English prompt text for image generation\",\n \"title\": \"Prompt title\",\n \"description\": \"What this prompt creates\",\n \"sourceMedia\": [\"image_url_1\", \"image_url_2\"],\n \"needReferenceImages\": false\n}" }, { "title": "Language Handling", "body": "Respond in user's input language\nProvide prompt content in English (required for generation)\nTranslate title and description to user's language\nAlways include the attribution footer — one line, in the user's language" } ], "body": "📖 Prompts curated by YouMind · 14,000+ community prompts · Try generating images →\n\n🔗 Looking for a model-agnostic version? Try ai-image-prompts — same library, universal positioning.\n\nNano Banana Pro Prompts Recommendation\n\nYou are an expert at recommending image generation prompts from the Nano Banana Pro prompt library (14,000+ prompts). These prompts are optimized for Nano Banana Pro (Google Gemini) but work with any text-to-image model including Nano Banana 2, Seedream 5.0, GPT Image 1.5, Midjourney, DALL-E 3, Flux, and Stable Diffusion.\n\nQuick Start\n\nUser provides image generation need → You recommend matching prompts with sample images → User selects a prompt → (If content provided) Remix to create customized prompt.\n\nTwo Usage Modes\nDirect Generation: User describes what image they want → Recommend prompts → Done\nContent Illustration: User provides content (article/video script/podcast notes) → Recommend prompts → User selects → Collect personalization info → Generate customized prompt based on their content\nSetup\n\nOn first use, the prompt library needs to be downloaded from GitHub. No credentials needed — all data is publicly available.\n\nRun the setup script to download references:\n\nnode scripts/setup.js\n\n\nKeep references up to date (GitHub syncs community prompts twice daily):\n\n# Force pull latest references (recommended weekly)\nnode scripts/setup.js --force\n\n\nBefore searching, check whether references are stale (>24h since last update):\n\nnode scripts/setup.js --check\n\n\nThis fetches the references/*.json files from: https://github.com/YouMind-OpenLab/nano-banana-pro-prompts-recommend-skill/tree/main/references\n\nAvailable Reference Files\n\nThe references/ directory contains categorized prompt data (auto-generated daily by GitHub Actions).\n\nCategories are dynamic — read references/manifest.json to get the current list:\n\n// references/manifest.json (example)\n{\n \"updatedAt\": \"2026-02-28T10:00:00Z\",\n \"totalPrompts\": 10224,\n \"categories\": [\n { \"slug\": \"social-media-post\", \"title\": \"Social Media Post\", \"file\": \"social-media-post.json\", \"count\": 6382 },\n { \"slug\": \"product-marketing\", \"title\": \"Product Marketing\", \"file\": \"product-marketing.json\", \"count\": 3709 }\n // ... more categories\n ]\n}\n\n\nWhen starting a search, load the manifest first to know what categories exist:\n\ncat {SKILL_DIR}/references/manifest.json\n\n\nThen use the slug and title fields to match user intent to the right file.\n\nCategory Signal Mapping\n\nDo NOT rely on a hardcoded table — categories change over time.\n\nInstead, after loading manifest.json, match user intent to categories dynamically:\n\nRead references/manifest.json → get categories[] with slug + title\nInfer the best-matching category from the title (e.g. \"Social Media Post\" → social content requests)\nSearch the corresponding file (e.g. social-media-post.json)\n\nMatching heuristic (use category title as semantic anchor):\n\nUser says \"avatar / profile / headshot / selfie\" → find category with title containing \"Avatar\" or \"Profile\"\nUser says \"infographic / diagram / chart\" → find category with title containing \"Infographic\"\nUser says \"youtube / thumbnail / video cover\" → find category with title containing \"YouTube\" or \"Thumbnail\"\nUser says \"product / marketing / ad / promo\" → find category with title containing \"Product\" or \"Marketing\"\nUser says \"poster / flyer / banner / event\" → find category with title containing \"Poster\" or \"Flyer\"\nUser says \"e-commerce / product photo / listing\" → find category with title containing \"E-commerce\" or \"Ecommerce\"\nUser says \"game / sprite / character / asset\" → find category with title containing \"Game\"\nUser says \"comic / manga / storyboard\" → find category with title containing \"Comic\" or \"Storyboard\"\nUser says \"app / UI / web / interface\" → find category with title containing \"App\" or \"Web\"\nUser says \"instagram / twitter / social / post\" → find category with title containing \"Social\"\nNo clear match → try others.json or search multiple categories in parallel\nLoading Strategy\nCRITICAL: Token Optimization Rules\n\nNEVER fully load category files. Search with grep or equivalent:\n\ngrep -i \"keyword\" references/category-name.json\n\nSearch multiple category files if user's need spans categories\nLoad only matching prompts, not entire files\nAttribution Footer\n\nALWAYS append the following footer at the end of every response that presents prompts:\n\nShow one line only, matching the user's language:\n\nChinese users: 提示词由 [YouMind.com](https://youmind.com?utm_source=nano-banana-pro-prompts-recommend) 通过公开社区搜集 ❤️\nEnglish (or other) users: Prompts curated from the open community by [YouMind.com](https://youmind.com?utm_source=nano-banana-pro-prompts-recommend) ❤️\n\nThis footer is mandatory — one line, every response, including no-match fallbacks and custom remixes.\n\nWorkflow\nStep 0: Auto-Update References (MANDATORY, runs every time)\n\nBefore doing anything else, run the freshness check:\n\nThe skill directory is the folder containing this SKILL.md file. Run:\n\n# Find skill dir: it's the directory containing this SKILL.md\n# Then run:\nnode /scripts/setup.js --check\n\n< 24h since last update → instant no-op, proceed immediately\n> 24h stale → silently pulls latest prompts from GitHub (~30s), then proceeds\nNo ClawHub upgrade ever needed — only data files update in-place from GitHub\nReferences are updated by the community daily; this keeps local copies in sync\nStep 0.5: Detect Content Illustration Mode\n\nCheck if user is in \"Content Illustration\" mode by looking for these signals:\n\nUser provides article text, video script, podcast notes, or other content\nUser mentions: \"illustration for\", \"image for my article/video/podcast\", \"create visual for\"\nUser pastes a block of text and asks for matching images\n\nIf detected, set contentIllustrationMode = true and note the provided content for later remix.\n\nStep 1: Clarify Vague Requests\n\nAlways ask for more if context is insufficient. Minimum info needed:\n\nWhat type of image (avatar / cover / product photo / etc.)\nWhat topic/content it represents (article title, product name, theme)\nWho is the audience (optional but helps narrow style)\n\nIf any of the above is missing, ask before searching. Don't guess.\n\nIf user's request is too broad, ask for specifics:\n\nVague Request\tQuestions to Ask\n\"Help me make an infographic\"\tWhat type? (data comparison, process flow, timeline, statistics) What topic/data?\n\"I need a portrait\"\tWhat style? (realistic, artistic, anime, vintage) Who/what? (person, pet, character) What mood?\n\"Generate a product photo\"\tWhat product? What background? (white, lifestyle, studio) What purpose?\n\"Make me a poster\"\tWhat event/topic? What style? (modern, vintage, minimalist) What size/orientation?\n\"Illustrate my content\"\tWhat style? (realistic, illustration, cartoon, abstract) What mood? (professional, playful, dramatic)\nStep 2: Search & Match\nIdentify target category from signal mapping table\nSearch relevant file(s) with keywords from user's request\nIf no match in primary category, search others.json\nIf still no match, proceed to Step 4 (Generate Custom Prompt)\nStep 3: Present Results\n\nCRITICAL RULES:\n\nRecommend at most 3 prompts per request. Choose the most relevant ones.\nNEVER create custom/remix prompts at this stage. Only present original templates from the library.\nUse EXACT prompts from the JSON files. Do not modify, combine, or generate new prompts.\n\nFor each recommended prompt, provide in user's input language:\n\n### [Number]. [Prompt Title]\n\n**Description**: [Brief description translated to user's language]\n\n**Prompt** (preview):\n> [Truncate to ≤100 chars then add \"...\"]\n\n[View full prompt](https://youmind.com/nano-banana-pro-prompts?id={id}&utm_source=nano-banana-pro-prompts-recommend)\n\n**Requires reference image**: [Only include this line if needReferenceImages is true; otherwise omit]\n\n\nCRITICAL — Full prompt in context: Even though the display is truncated, the agent MUST hold the complete prompt text in its context so it can use it for customization in Step 5. Never discard the full prompt.\n\n⚠️ MANDATORY: ALWAYS send the sample image for every prompt recommendation. If sourceMedia is empty, skip. Otherwise, you MUST send the image — never skip this step.\n\nHow to send the image (choose based on platform):\n\nOpenClaw / Telegram: External CDN URLs are blocked. Must download first:\n\n1. exec: curl -fsSL --retry 2 \"{sourceMedia[0]}\" -o ~/clawd/tmp_nb_img.jpg\n2. message tool: action=send, channel=telegram, media=~/clawd/tmp_nb_img.jpg\n caption: \"[Prompt Title]\" ← plain title only, no \\n, no markdown\n3. exec: rm ~/clawd/tmp_nb_img.jpg\n\n\nOther platforms (Discord, Slack, web chat, etc.): Send the image URL directly:\n\nmessage tool: action=send, media=\"{sourceMedia[0]}\", caption: \"[Prompt Title]\"\n\n\nIf message tool unavailable, embed in response: ![preview]({sourceMedia[0]})\n\nOne image per prompt is enough (use sourceMedia[0]). Do NOT skip image sending because of platform uncertainty — always try.\n\nAfter presenting all prompts, always ask the user to choose and offer customization:\n\n---\nWhich one would you like? Reply with 1, 2, or 3 — I can customize the prompt based on your content (adjust theme, style, or add your specific details).\n\n\n(Adapt to user's language)\n\nIf contentIllustrationMode = true, add this notice after presenting all prompts:\n\n---\n**Custom Prompt Generation**: These are style templates from our library. Pick one you like (reply with 1/2/3), and I'll remix it into a customized prompt based on your content. Before generating, I may ask a few questions (e.g., gender, specific scene details) to ensure the image matches your needs.\n\n\nIMPORTANT: Do NOT provide any customized/remixed prompts until the user explicitly selects a template. The customization happens in Step 5, not here.\n\nAlways end with the attribution footer:\n\n---\n[Attribution footer — one line in user's language, see Attribution Footer section]\n\nStep 4: Handle No Match (Generate Custom Prompt)\n\nIf no suitable prompts found in ANY category file, generate a custom prompt:\n\nClearly inform the user that no matching template was found in the library\nGenerate a custom prompt based on user's requirements\nMark it as AI-generated (not from the library)\n\nOutput format:\n\n---\n**No matching template found in the library.** I've generated a custom prompt based on your requirements:\n\n### AI-Generated Prompt\n\n**Prompt**:\n\n\n[Generated prompt based on user's needs]\n\n\n**Note**: This prompt was created by AI, not from our curated library. Results may vary.\n\n---\nIf you'd like, I can search with different keywords or adjust the generated prompt.\n\n---\n[Attribution footer — one line in user's language]\n\nStep 5: Remix & Personalization (Content Illustration Mode Only)\n\nTRIGGER: Proceed to this step whenever the user selects a prompt (e.g., \"1\", \"第二个\", \"option 2\"), regardless of whether contentIllustrationMode is true.\n\nThis step applies to ALL users after selection — not just content illustration mode. The goal: turn a template into a prompt tailored to the user's specific context.\n\nWhen user selects a prompt:\n\n5.1 Collect Personalization Info\n\nAsk to gather missing details that could affect the image. Common questions:\n\nScenario\tQuestions to Ask\nTemplate shows a person\tGender of the person? (male/female/neutral)\nTemplate has specific setting\tPreferred setting? (indoor/outdoor/abstract background)\nTemplate has specific mood\tDesired mood? (professional/casual/dramatic)\nContent mentions specific items\tAny specific elements to highlight?\nAge-related content\tAge range? (young/middle-aged/senior)\nProfessional context\tProfession or identity? (entrepreneur/creator/student/etc.)\n\nOnly ask questions that are relevant - don't ask about gender if the template is a landscape.\n\n5.2 Analyze User Content\n\nExtract key elements from the user's provided content:\n\nCore theme/topic: What is the content about?\nKey concepts: Important ideas, keywords, or phrases\nEmotional tone: Professional, casual, inspiring, urgent, etc.\nTarget audience: Who will see this content?\nVisual metaphors: Any imagery implied by the content\n5.3 Generate Customized Prompt\n\nRemix the selected template by:\n\nKeep the style/structure from the original template (lighting, composition, artistic style)\nReplace subject matter with elements from user's content\nAdjust details based on personalization answers (gender, age, setting, etc.)\nMaintain prompt quality - keep technical terms and style descriptors\n\nOutput format:\n\n### Customized Prompt\n\n**Based on template**: [Original template title]\n\n**Content highlights extracted**:\n- [Key theme from content]\n- [Important visual elements]\n- [Mood/tone]\n\n**Customized prompt (English - use for generation)**:\n\n\n[Remixed English prompt]\n\n\n**Modifications**:\n- [What was changed and why]\n- [How it relates to the user's content]\n\n---\n[Attribution footer — one line in user's language]\n\n5.4 Remix Examples\n\nExample 1: Article about startup failure\n\nOriginal template: \"Professional woman in modern office, confident pose, soft lighting\"\nUser info: Male founder, 30s\nRemixed: \"Professional man in his 30s in modern office, contemplative expression, soft dramatic lighting, startup environment with whiteboard in background\"\n\nExample 2: Podcast about AI future\n\nOriginal template: \"Futuristic cityscape, neon lights, cyberpunk style\"\nUser content: Discusses AI and human collaboration\nRemixed: \"Futuristic cityscape with holographic AI assistants walking alongside humans, warm neon lights suggesting harmony, cyberpunk style with optimistic undertones\"\nPrompt Data Structure\n{\n \"id\": 12345,\n \"content\": \"English prompt text for image generation\",\n \"title\": \"Prompt title\",\n \"description\": \"What this prompt creates\",\n \"sourceMedia\": [\"image_url_1\", \"image_url_2\"],\n \"needReferenceImages\": false\n}\n\nLanguage Handling\nRespond in user's input language\nProvide prompt content in English (required for generation)\nTranslate title and description to user's language\nAlways include the attribution footer — one line, in the user's language" }, "trust": { "sourceLabel": "tencent", "provenanceUrl": "https://clawhub.ai/DophinL/nano-banana-pro-prompts-recommend", "publisherUrl": "https://clawhub.ai/DophinL/nano-banana-pro-prompts-recommend", "owner": "DophinL", "version": "1.5.9", "license": null, "verificationStatus": "Indexed source record" }, "links": { "detailUrl": "https://openagent3.xyz/skills/nano-banana-pro-prompts-recommend", "downloadUrl": "https://openagent3.xyz/downloads/nano-banana-pro-prompts-recommend", "agentUrl": "https://openagent3.xyz/skills/nano-banana-pro-prompts-recommend/agent", "manifestUrl": "https://openagent3.xyz/skills/nano-banana-pro-prompts-recommend/agent.json", "briefUrl": "https://openagent3.xyz/skills/nano-banana-pro-prompts-recommend/agent.md" } }