<div data-type="blurb" data-id="94">Drop in my <a target="_blank" rel="ugc" href="https://t.me/+NIc-ia0Z_BpkYmJh">Telegram Group</a> and <a target="_blank" rel="ugc" href="https://discord.gg/kF7mPKFtRt">Discord group</a> for chats, questions, feedbacks. 欢迎加入<a target="_blank" rel="ugc" href="https://t.me/+NIc-ia0Z_BpkYmJh">纸飞机</a>和<a target="_blank" rel="ugc" href="https://discord.gg/kF7mPKFtRt">Discord</a>群组。RH体验地址（邀请码Invite Code: cptpovgo）：<a target="_blank" rel="ugc" href="https://www.runninghub.ai/user-center/1993809672580788226?inviteCode=cptpovgo">https://www.runninghub.ai/user-center/1993809672580788226?inviteCode=cptpovgo</a> 喜欢我的作品吗？🎨 我会持续免费分享更多优质内容✨，欢迎支持一下，请我喝杯咖啡吧！💖 Like my work? 🎨 I share most of my work for free ✨, so support me by buying me a coffee ☕️! Drop by my discord channel for supporter perks. 想赞助一下？来我 Discord，里面有链接和特别福利和专属模型下载哦！CivitAI 邀请码 Referral Code: R44VYP99 Extra 2.5k Buzz on membership purchase， 买会员送2千5闪电!Ko-fi: <a target="_blank" rel="ugc" href="http://ko-fi.com/catlover1937">ko-fi.com/catlover1937</a>Please do not redistribute without consent. 如非作者同意禁止搬运到其他平台，如国内Up主需要转发网盘请备注出处。<h1>WOULD YOU PLEASE CLICK THE LIKE BUTTON? 拜托动动您可爱的小手点个赞好吗？</h1></div><hr /><h2 id="moody-moss-prompt-skill-release">🎉 Moody MOSS Prompt Skill Release</h2>I'm releasing my personal MOSS-Image prompting skill, specifically tailored for Moody MOSS checkpoints.When MOSS isn't following your prompt, it usually comes down to two things:<ol><li>🚫 Model-side limitations or restrictions are affecting the generation.</li><li>✍️ Your prompt writing sucks.</li></ol>And if you're already using my model, chances are 90% of the time it's probably #2. 😅People constantly say:❌ "The model doesn't work." ❌ "It doesn't follow my prompt." ❌ "MOSS completely ignores what I write."But most of the time, the model is perfectly capable — your prompt just isn't communicating your ideas clearly enough.<h3 id="stop-hand-writing-your-prompts.">🛑 Stop hand-writing your prompts.</h3>Humans write prompts the way humans think. But AI doesn't necessarily interpret language the same way we do.Instead of spending hours figuring out the perfect wording, why not let AI handle that part?Feed Moody MOSS Prompt Skill into your LLM / AI Agent, describe what you want in normal language, and let AI translate your ideas into prompts that MOSS can understand more effectively. 🤖✨<h3 id="more-than-just-a-prompt-enhancer">🧠 More Than Just a Prompt Enhancer</h3>Unlike a basic prompt generator, this Skill supports two different prompting approaches:📸 Caption Mode — Perfect for photography, portraits, cinematic scenes, illustrations, and everyday image generation. It transforms simple ideas into detailed, coherent visual descriptions, covering composition, lighting, camera angles, materials, atmosphere, and more.📐 Layout Mode — Designed for more complex compositions that require precise spatial control, including posters, typography, multiple subjects, object placement, and structured visual layouts. It uses bounding boxes to help define where elements should appear in the image.The Skill automatically selects the appropriate approach based on your request, while preserving your original creative intent.It also includes a prompt compilation workflow for converting structured Layout drafts into MOSS-compatible prompts.All of this is built around the prompting structure, techniques, and practical experience I use myself, specifically adapted for Moody MOSS models.AI knows how to communicate with AI better than most humans do.You focus on the creativity. Let AI handle the technical prompting.💡 Before blaming the model, fix the prompt.<h3 id="special-thanks">❤️ Special Thanks</h3>Special thanks to technical support and sharing from @salttaro .He is an absolute master of AIGC with insane technical backround.You can find him in the Moody Discord channel. If you care about cinematic prompting, camera language, composition, storytelling, and actually getting AI to understand your scene, definitely check out his work.<h2 id="moody-moss-prompt-skill">🎉 Moody MOSS Prompt Skill 正式发布</h2>这次我把自己使用的 MOSS-Image 提示词 Skill 也放出来了，专门针对 Moody MOSS 系列 Checkpoint 做了适配和优化。当你发现 MOSS 不怎么听提示词的时候，通常就两个原因：<ol><li>🚫 模型本身的限制影响了生成结果</li><li>✍️ 你的提示词写得不行</li></ol>如果你已经在用我的模型，那说实话，90% 的情况大概率是第二种。 😅我经常看到有人说：❌ “这个模型不行。” ❌ “完全不听提示词。” ❌ “MOSS 根本不理解我写的东西。”但很多时候，不是模型不会，而是你的提示词根本没有把你的想法准确地传达给模型。<h3>🛑 所以，别再手搓提示词了。</h3>人类写提示词，是按照人类自己的语言习惯和思维方式去写的。但 AI 理解语言的方式，并不完全和人一样。与其花大量时间研究怎么组织语言，为什么不直接让 AI 帮你完成这部分工作呢？你只需要把 Moody MOSS Prompt Skill 喂给你的 LLM / AI Agent，然后用正常语言告诉它你想要什么，让 AI 帮你把想法转换成更适合 MOSS 理解的提示词。🤖✨<h3>🧠 不只是普通的提示词优化工具</h3>和简单的提示词生成器不同，这个 Skill 支持 两种不同的提示词生成模式：📸 Caption Mode（自然语言描述模式）适合人物摄影、写真、电影感场景、插画，以及各种日常生图需求。它可以把你的简单想法扩展成完整、连贯的视觉描述，包括人物动作、构图、镜头角度、光线、材质、环境氛围等，让模型更容易理解你真正想生成的画面。📐 Layout Mode（空间布局模式）适合需要精确控制画面元素位置的复杂场景，例如海报设计、文字排版、多人物构图、物体摆放，以及各种需要明确空间关系的图像。通过 Bounding Box（边界框）定义画面元素的位置和范围，让模型获得更加明确的空间布局信息。Skill 会根据你的需求自动判断更适合使用哪种模式，同时尽可能保留你原本的创作意图。此外，它还包含了一套提示词编译流程，可以把结构化的 Layout 草稿转换成 MOSS 可以直接使用的提示词格式。整个 Skill 融合了我自己平时使用的提示词结构、表达方式和实践经验，并且专门针对 Moody MOSS 模型 做了调整。让 AI 去跟 AI 沟通，很多时候比人类自己硬写效果更好。你负责发挥创意，剩下的技术性提示词工作交给 AI 就行。💡 所以在怪模型之前，先看看是不是你的提示词有问题。<h3>❤️ 特别感谢</h3>特别感谢芋头哥的技术支持和分享 @salttaro 。他是真的技术大牛。🔥你可以在 Moody Discord 频道 找到他。如果你对电影感提示词、镜头语言、构图、场景调度、叙事，以及怎么让 AI 真正理解你脑子里的画面感兴趣，强烈建议去看看他的作品。

Moody Moss Prompt Writing Skill - LLM SYSTEM PROMPT

Files with SHA256: d9b33b63b3767503c07e81a5d6ab7d0a3005e3ce1e38feb47c1ce9553295b639