Skip to content

pigeonai-yang/oneirloom

v0.3.1-previewMIT

Visual prompting, reference analysis, craft construction, digital form, spatial concepts, and result diagnosis with 23 source skills.

oneirloom

Oneirloom (织梦师), also called 织梦师Skill or Oneirloom Skill. Explore visual directions across media, then turn visual requests and references into complete image prompts, graphic designs, craft constructions, digital forms, spatial concepts, sticker packs, icon assets, or brand identities. Diagnose results, route illustrated tutorial writing, or package an image and prompt into a 3:4 card. Select only the methods and templates the task needs.

oneirloom-brand-identity

Design a coherent brand visual identity across marks, typography, color, graphic language, real applications, and a usable guideline book. Use for a full VI system or a defined identity extension; distinguish a concept brand board from delivered master assets and applications.

oneirloom-camera-composition

将景别、机位、视角、透视、遮挡、裁切和运动意图写成明确的画面空间关系。用于摄影、电影感、广角仰拍、俯拍、主体布局和参考图构图复现。

oneirloom-character-sheet

Design character sheets, wardrobe sheets, multi-view identity anchors, and production or expression cards. Select indexed wardrobe, four-view plus four-expression, or three-view templates when their layout fits; ordinary single-image reconstruction belongs to visual analysis.

oneirloom-color-light

把色彩、明度、对比、光向、材质受光和空气感转成区域明确的图像提示词。用于修正偏灰、偏冷、过曝、光色错位或“鲜活梦幻”等抽象氛围。

oneirloom-craft-construction

Turn a visual subject into a clear craft construction when units, joins, thickness, edges, or support define the result. Use for work such as fuse-bead art or layered paper relief.

oneirloom-digital-form

Shape or transform a digital visual subject when silhouette, volume, repetition, deformation, connections, or composite layers define the result. Use for digital-native forms and 3D concept renders.

oneirloom-dream-weaving

Develop vague feelings, memories, fragmentary imagery, or an unsatisfying concept into a drawable visual scene through collaborative conversation. Preserve the initial feeling and user choices; hand clear briefs directly to the existing production capability.

oneirloom-expression-stickers

Design chat reaction stickers and emote packs with recognizable character identity, distinct communicative intent, gestures, and optional exact captions. Use for individual stickers or coordinated packs, not character turnarounds, UI icons, or brand logos.

oneirloom-figure-art

Shape non-explicit artistic prompts for clearly adult unclothed figures, including requested concealment; coordinate visible anatomy and view-dependent coverage while leaving medium and style to their own branches.

oneirloom-icon-design

Design coherent functional UI icon families and distinctive app or product symbols. Adapt existing icon languages, create editable vector assets or requested illustrative icon prompts, and inspect meaning and legibility at the intended size. Reaction stickers use oneirloom-expression-stickers.

oneirloom-image-tutorial

Write, expand, or polish illustrated image-generation tutorials with concrete prompt examples and inspected results. Use for teaching articles, section-by-section expansion, and tutorial copy editing; not for a standalone image prompt or reference-image reconstruction.

oneirloom-model-krea-2

为 Krea 2 Medium、Large、Turbo 及开放权重 RAW/Turbo 编写和校验提示词,区分托管网站、API 与自部署入口的参考图和参数能力。用于 Krea 2 文生图、风格参考和跨模型迁移。

oneirloom-model-qwen-image-2-1

为 Qwen-Image-2.1 文生图、图像编辑、透明背景与多参考图任务编写提示词,并核对当前部署入口是否支持相应输入。用于 Qwen-Image-2.1 提示词生成和跨模型迁移。

oneirloom-model-qwen-image-2-1-comfyui-edit

在本地 ComfyUI 上编排 Qwen-Image-2.1 图像编辑流程:先判断需求属于真编辑还是重绘,据此接线、设置 denoise、控制画幅、分配多参考图角色,按官方 I2I 规则写编辑提示词,并在交付前验证改动范围、画幅与身份一致性。

oneirloom-product-art-direction

Inspect supplied product references and develop static ecommerce images, realistic product scenes, or creative advertisements. Resolve clear or vague briefs, preserve product facts through permitted transformations, and revise complete prompts across turns.

oneirloom-prompt-card

Create a 3:4 vertical card from an existing image and complete prompt. Crop the image to fill the upper third, and typeset the prompt in the lower two thirds. Use for image-and-prompt collages, prompt cards, or Oneirloom prompt sharing layouts.

oneirloom-result-diagnosis

Compare an actual or user-reported result miss with the source, current intent, and submitted prompt. Separate source-to-prompt from prompt-to-result errors and return a complete correction that preserves successful relations. Ordinary requested edits need no diagnosis.

oneirloom-space-conception

Organize a room, set, or environment when masses, enclosure, openings, adjacency, circulation, focus, scale, or depth define the visual result. Use before selecting a camera view.

oneirloom-style-design

Understand visual intent and subject identity, observe references, and design composed graphics before choosing production. Write complete prompts or create requested posters, covers, editorial layouts, infographics, composed ads, and packaging graphics. Use product art direction for supplied-product ads.

oneirloom-style-illustration

Compose watercolor, print, comic, animation, concept-art, folkloric narrative, and paper-collage prompts through visible linework, color layers, texture, and spatial structure. Select indexed presets for decorative rainy-night prints, halftone portrait cutouts, or zhiguai narrative illustrations.

oneirloom-style-photography

Compose realistic photography, lifestyle and fashion portraits, product photographs, and film-still prompts. Select indexed scene templates for requests such as crouching selfies or indoor full-length portraits.

oneirloom-visual-analysis

拆解参考图或已成形画面的模糊视觉需求,提取可检验的主体、空间、区域色彩和媒介锚点。用于参考图反推、图像复现、跨模型迁移与将“梦幻”“高级”等抽象词转成具体画面关系。