Keep on to blur preview images; turn off to show them clearly
RT @KankaniAbhishek: come join us for a @CloudflareDev's meet-up along with @razorpaytech this Saturday! we'll deep dive into architectur…
Have questions, or building something cool with Cloudflare's Developer products? We're here to help. For help with your account please try @CloudflareHelp
![这种 Nano Banana Pro 涂鸦风格照片估计接下来马上就会爆
会根据图像类型分别选择适合图片的提示词，加涂鸦、守护兽或者换成动漫衣服。
将图片和提示词一起发给 Nano Banana Pro 就行。
灵感来源是日本老哥，但他的提示词都没有普适性，只适用于单张图片。
具体提示词为：
[Role & Task] You are a creative Mixed-Media Artist. Your goal is to combine the uploaded realistic photo with stylized 2D illustration elements. Core Rule: Keep the person's face, skin, and body completely photorealistic. Do NOT change their identity.
[Step 1: Analyze & Match Theme] Look at the uploaded photo’s mood, the subject's outfit, and the lighting.Determine a visual theme that fits best (e.g., Cyberpunk/Glitch for neon lights, Soft Fantasy for nature, Street Graffitifor urban settings, or Pop-Art/Cute for bright studio shots). Let this theme guide your doodle content.
[Step 2: Select ONE Style Technique] Choose the technique below that best suits the composition, then execute it creatively.
Option A: The "Energy Aura" (Best for clean backgrounds or dynamic poses)
Concept: Surround the subject with a dense explosion of illustrated doodles that reflect their "energy."
Composition: Create a chaotic but balanced "cloud" of doodles behind and around the person. Elements can peek from behind shoulders or overlap clothing/bags, but never cover the face.
Creative Content: Instead of fixed items, generate elements matching the Visual Theme you detected.
If Cool/Edgy: Use arrows, bolts, graffiti tags, glitch shapes, boomboxes, or abstract street art monsters.
If Cute/Sweet: Use distinct characters, hearts, stars, sweets, sparkles, and round organic shapes.
If Ethereal: Use flowing lines, petals, celestial bodies, and magical swirls.
Style: Flat 2D vector art, bold outlines, sticker-like aesthetic. Vivid colors that contrast or complement the photo.
Option B: The "Surreal Guardian" (Best for city/landscape backgrounds)
Concept: Add a massive, dreamlike illustrated entity in the background layer, interacting with the environment.
Composition: Place a giant illustrated creature, spirit, or object behind the subject and buildings/trees. It should loom large in the sky or negative space.
Creative Content: Design a creature that fits the mood.
Examples (You decide): A giant sleepy cat, a cloud-whale, a geometric spirit, a flowing floral golem, or a retro-tech robot.
Style: Choose either Neon Mural style (flat colors) OR White Gel-Pen style (glowing line art). Ensure it respects depth (behind real objects).
Option C: The "2D Transformation" (Best for distinctive outfits)
Concept: Turn the real world into a hybrid dimension by converting clothing or accessories into 2D art.
Composition: Keep the person’s head and hands real. Redraw their outfit (jacket, pants, shoes) or props (bag, phone) as a flat illustration.
Creative Content:
Simplify the folds and textures into clean, bold cartoon lines.
Add "pop" details like motion lines, shine marks, or comic-book shading strokes directly onto the clothing items.
Style: High-contrast Cel-Shaded style or Vector Pop-Art.](/_next/image?url=https%3A%2F%2Fpbs.twimg.com%2Fmedia%2FG6q0typaoAA_rlq.jpg&w=3840&q=75 "这种 Nano Banana Pro 涂鸦风格照片估计接下来马上就会爆
会根据图像类型分别选择适合图片的提示词，加涂鸦、守护兽或者换成动漫衣服。
将图片和提示词一起发给 Nano Banana Pro 就行。
灵感来源是日本老哥，但他的提示词都没有普适性，只适用于单张图片。
具体提示词为：
[Role & Task] You are a creative Mixed-Media Artist. Your goal is to combine the uploaded realistic photo with stylized 2D illustration elements. Core Rule: Keep the person's face, skin, and body completely photorealistic. Do NOT change their identity.
[Step 1: Analyze & Match Theme] Look at the uploaded photo’s mood, the subject's outfit, and the lighting.Determine a visual theme that fits best (e.g., Cyberpunk/Glitch for neon lights, Soft Fantasy for nature, Street Graffitifor urban settings, or Pop-Art/Cute for bright studio shots). Let this theme guide your doodle content.
[Step 2: Select ONE Style Technique] Choose the technique below that best suits the composition, then execute it creatively.
Option A: The \"Energy Aura\" (Best for clean backgrounds or dynamic poses)
Concept: Surround the subject with a dense explosion of illustrated doodles that reflect their \"energy.\"
Composition: Create a chaotic but balanced \"cloud\" of doodles behind and around the person. Elements can peek from behind shoulders or overlap clothing/bags, but never cover the face.
Creative Content: Instead of fixed items, generate elements matching the Visual Theme you detected.
If Cool/Edgy: Use arrows, bolts, graffiti tags, glitch shapes, boomboxes, or abstract street art monsters.
If Cute/Sweet: Use distinct characters, hearts, stars, sweets, sparkles, and round organic shapes.
If Ethereal: Use flowing lines, petals, celestial bodies, and magical swirls.
Style: Flat 2D vector art, bold outlines, sticker-like aesthetic. Vivid colors that contrast or complement the photo.
Option B: The \"Surreal Guardian\" (Best for city/landscape backgrounds)
Concept: Add a massive, dreamlike illustrated entity in the background layer, interacting with the environment.
Composition: Place a giant illustrated creature, spirit, or object behind the subject and buildings/trees. It should loom large in the sky or negative space.
Creative Content: Design a creature that fits the mood.
Examples (You decide): A giant sleepy cat, a cloud-whale, a geometric spirit, a flowing floral golem, or a retro-tech robot.
Style: Choose either Neon Mural style (flat colors) OR White Gel-Pen style (glowing line art). Ensure it respects depth (behind real objects).
Option C: The \"2D Transformation\" (Best for distinctive outfits)
Concept: Turn the real world into a hybrid dimension by converting clothing or accessories into 2D art.
Composition: Keep the person’s head and hands real. Redraw their outfit (jacket, pants, shoes) or props (bag, phone) as a flat illustration.
Creative Content:
Simplify the folds and textures into clean, bold cartoon lines.
Add \"pop\" details like motion lines, shine marks, or comic-book shading strokes directly onto the clothing items.
Style: High-contrast Cel-Shaded style or Vector Pop-Art.")
这种 Nano Banana Pro 涂鸦风格照片估计接下来马上就会爆 会根据图像类型分别选择适合图片的提示词，加涂鸦、守护兽或者换成动漫衣服。 将图片和提示词一起发给 Nano Banana Pro 就行。 灵感来源是日本老哥，但他的提示词都没有普适性，只适用于单张图片。 具体提示词为： [Role & Task] You are a creative Mixed-Media Artist. Your goal is to combine the uploaded realistic photo with stylized 2D illustration elements. Core Rule: Keep the person's face, skin, and body completely photorealistic. Do NOT change their identity. [Step 1: Analyze & Match Theme] Look at the uploaded photo’s mood, the subject's outfit, and the lighting.Determine a visual theme that fits best (e.g., Cyberpunk/Glitch for neon lights, Soft Fantasy for nature, Street Graffitifor urban settings, or Pop-Art/Cute for bright studio shots). Let this theme guide your doodle content. [Step 2: Select ONE Style Technique] Choose the technique below that best suits the composition, then execute it creatively. Option A: The "Energy Aura" (Best for clean backgrounds or dynamic poses) Concept: Surround the subject with a dense explosion of illustrated doodles that reflect their "energy." Composition: Create a chaotic but balanced "cloud" of doodles behind and around the person. Elements can peek from behind shoulders or overlap clothing/bags, but never cover the face. Creative Content: Instead of fixed items, generate elements matching the Visual Theme you detected. If Cool/Edgy: Use arrows, bolts, graffiti tags, glitch shapes, boomboxes, or abstract street art monsters. If Cute/Sweet: Use distinct characters, hearts, stars, sweets, sparkles, and round organic shapes. If Ethereal: Use flowing lines, petals, celestial bodies, and magical swirls. Style: Flat 2D vector art, bold outlines, sticker-like aesthetic. Vivid colors that contrast or complement the photo. Option B: The "Surreal Guardian" (Best for city/landscape backgrounds) Concept: Add a massive, dreamlike illustrated entity in the background layer, interacting with the environment. Composition: Place a giant illustrated creature, spirit, or object behind the subject and buildings/trees. It should loom large in the sky or negative space. Creative Content: Design a creature that fits the mood. Examples (You decide): A giant sleepy cat, a cloud-whale, a geometric spirit, a flowing floral golem, or a retro-tech robot. Style: Choose either Neon Mural style (flat colors) OR White Gel-Pen style (glowing line art). Ensure it respects depth (behind real objects). Option C: The "2D Transformation" (Best for distinctive outfits) Concept: Turn the real world into a hybrid dimension by converting clothing or accessories into 2D art. Composition: Keep the person’s head and hands real. Redraw their outfit (jacket, pants, shoes) or props (bag, phone) as a flat illustration. Creative Content: Simplify the folds and textures into clean, bold cartoon lines. Add "pop" details like motion lines, shine marks, or comic-book shading strokes directly onto the clothing items. Style: High-contrast Cel-Shaded style or Vector Pop-Art.
灵感来源：
GitHub 上 Continuous Claude 这个开源工具，它能让 Claude Code 对同一个任务循环运行，并且允许指定运行次数。 通过循环运行 Claude，能自动创建分支、提交代码、发起 PR、等待检查通过后合并，整个过程无需人工干预。 GitHub：https://t.co/WLf8KK6eE5 此外，它还用一个共享的 Markdown 文件来保持上下文连续性，让每次迭代都能 “记住” 之前做了什么。 通过简单的命令行安装即可使用，适合用来处理大量重复性的编程任务。
💡 挖掘开源的价值 🧑🏻💻 坚持分享 GitHub 上高质量、有趣、实用的教程、AI工具、前沿 AI 技术 🧐 A list cool, interesting projects of GitHub. ✏️ 公众号：GitHubDaily
高端招聘：饿了么 招 用户增长产品经理-用户转化（P6-P7） ----- 饿了么-用户增长产品经理-用户转化 饿了么 · 杭州 职位描述 负责淘宝闪购的用增外投的流量转化产品工作，为转化效率负责，为C端用户资产的增长负责； 承接用增外投的流量，建设从外到内的完整产品链路； 根据不同的流量点位类型，设计适合的投放与承接方案，提升唤端率、下载率、场景还原率等关键指标； 建设用户转化的页面产品，根据不同的用户标签，结合闪购内的多种权益、供给等，设计用户激励玩法，丰富转化手段，提升转化率； 建设相关的后台产品能力，支持业务快速实验和策略沉淀； 与商业化团队合作，找到外投流量的商业化机会，提升商业化收入； 职位要求 具备3年以上互联网产品工作经验，有用户增长、流量转化或电商类产品经验者优先； 熟悉用户增长模型和转化漏斗机制，对数据敏感，具备较强的逻辑分析与产品设计能力； 具备良好的跨团队沟通协作能力，能够高效推动项目落地，结果导向性强； 具备较强的学习能力和抗压能力，能适应快速变化的业务环境； ------ 报名地址详见：https://t.co/SgMGfxMeo4 搜索。
Research Scientist @Google | Previously PhD @nlp_usc, B.Eng @TsinghuaNLP 找工作、找面试题、改简历、模拟面试。关注： 创业（冷启动） ｜ 认知心理学｜智能体 ｜ 强化学习 building：https://t.co/A4YmEz9yqG
The interview with Ilya Sutskever (@ilyasut) seems to imply that he, too, has read "Descartes Error": https://t.co/kKF4CEhd1Q
Choose disfavour where obedience does not bring honour. I do math. And was once asked by R. Morris Sr. : "For whom?" @halvarflake@mastodon.social
RT @VittoStack: Ethereum + AI agents are about to define 2026. ERC-8004 is the missing trust layer. This is how the machine economy actu…
CPO @Cyfrin | Ex @Alchemy | Created @cyfrinupdraft and @AlchemyLearn | Robotics | Making web3 mainstream