[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M799THBR3XAX2DJCAGK3SPQK":3,"skill-same:01M799THBR3XAX2DJCAGK3SPQK":108},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":24,"commitSha":25,"skillDirectory":26,"forksCount":27,"repositoryHtmlUrl":28,"repositoryHomepageUrl":29,"ownerHtmlUrl":30,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":31,"install":32,"resources":45},"01M799THBR3XAX2DJCAGK3SPQK","studio-image",null,"Create and refine still images with Open Imagine's Image API. Direct composition, camera angle, framing, lens and lighting effects across photography, historical art, fantasy, anime, and illustration. Also covers images that carry text or layout (posters, covers, logos, app screens, infographics, packaging, signs, memes), characters and game assets, and reference edits such as restoration, composites and try-on. Use for shoots, coherent image sets, visual exploration, and reference-based edits with visual review.","CJHwong\u002Fopen-imagine","CJHwong","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F906057?v=4","01M799TF92MJNY7NAFGJ3222CY","Hoss","01M799TF9TG5QCDHK5TF50G9KJ",0,"MIT","https:\u002F\u002Fgithub.com\u002FCJHwong\u002Fopen-imagine\u002Fblob\u002Fmain\u002Fplugins\u002Fopen-imagine\u002Fskills\u002Fstudio-image\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: studio-image\ndescription: Create and refine still images with Open Imagine's Image API. Direct composition, camera angle, framing, lens and lighting effects across photography, historical art, fantasy, anime, and illustration. Also covers images that carry text or layout (posters, covers, logos, app screens, infographics, packaging, signs, memes), characters and game assets, and reference edits such as restoration, composites and try-on. Use for shoots, coherent image sets, visual exploration, and reference-based edits with visual review.\n---\n\n# Studio Image\n\nTurn a brief into still images that work together, then inspect and refine the actual images.\nOperate the user's Open Imagine server through its public API. Keep this workflow usable with\ndifferent Image models; discover their capabilities rather than assuming a particular checkpoint.\nRequires a reachable server and tools for HTTP, saving images, and visual inspection.\n\n## Choose the path\n\nIf the request is short or vague (\"a nice photo of my dog\"), or the user says they do not know where\nto start, follow the [brief-to-final flow](references\u002Fbrief-to-final.md). It asks at most three\nplain questions, shows four concepts from the [style library](references\u002Fstyle-library.md), and\nmakes the final from the concept the user picks. A user with a full brief (subject, use, style)\ncontinues below.\n\n## Establish the shoot\n\nUse the user's server URL; otherwise try `http:\u002F\u002F127.0.0.1:8765`. Read [the API operating\nnotes](references\u002Fstudio-api.md) before submitting work. If the server is unavailable, report the\nconnection problem; starting or installing a server is a separate task unless already authorized.\nOnce the backend is known, read [model notes](references\u002Fmodel-notes.md) for measured behavior on\nthat checkpoint. An unlisted model is untested; use its live capabilities and its own prompt\nwording.\n\nExtract the subject, intended use, deliverable count, aspect ratios, and visual direction from the\nbrief. Identify what must remain exact: a face, product silhouette, label, material, garment, or\nsupplied composition. Use sensible defaults for missing creative details; ask only when an unknown\nchanges the requested result materially. Respect the user's generation\u002Ftime budget. A request for\none finished image does not imply permission for an extensive concept round (the brief-to-final\nflow's four small concepts are its normal start, not an extra batch).\n\nGive each reference a role: subject identity, product geometry, wardrobe, lighting, pose, or layout.\nRefer to images in their submitted order using the selected model's supported wording. Distinguish a\nreference's required content from incidental background details. If a supplied product or person\nmust remain recognizable, prefer reference-based Edit over reinventing it from text. Do not claim\nexact brand or identity fidelity without inspecting the result.\n\nBefore you write the prompt, check the request against [boundaries](references\u002Fboundaries.md) if it\ninvolves a real person, a document, someone else's work, a brand or a mark, health, or news. Check\nit before Rewrite, because Rewrite carries the request into a longer prompt.\n\nAsk for the end use (screen, print, sticker, platform, vector, cutout) when the brief does not imply\nit, and otherwise state the end use you assume, because it sets the background, the margins and the\nsize; [delivery](references\u002Fdelivery.md) has the recipes. Split the job: the model makes the\npicture, and a layout tool or script adds what must be exact (long copy, exact colors, QR codes,\ngrids, counts, cut lines, the final pixel size, transparency).\n\nPlace each request on one accuracy level, and tell the user which:\n\n- **Mood only.** Mood, style and composition. The model is fine.\n- **Roughly right.** History, botany, products, architecture: it looks right and needs a human check\n  before anyone relies on it.\n- **Must be exact.** Data charts, maps with real geography, anatomy for teaching, floor plans, math\n  and chemistry figures, sewing patterns. Draw these with code or CAD, not this model.\n\n## Direct and generate\n\nFind the brief's kind of image in [use cases](references\u002Fuse-cases.md). It names the file to read, a\ndefault ratio, and whether Rewrite helped on the tested model. Before you write a custom prompt or\nuse Rewrite, read [prompt craft](references\u002Fprompt-craft.md). For an image that carries words or a\nlayout, read [design and text](references\u002Fdesign-and-text.md). For stickers, icons, characters,\ncomics, sprites or other drawn assets, read [illustration and\nassets](references\u002Fillustration-and-assets.md). For any Edit, read [editing](references\u002Fediting.md).\n\nFor composition, viewpoint, framing, focus, camera effects, lighting, or subject staging, read [shot\ndesign](references\u002Fshot-design.md). Treat these as independent decisions from style. Choose the\nvisible outcome first, then use compatible camera shorthand to support it. Camera numbers in prompt\ntext are creative cues, not calibrated API controls or proof of an actual capture setup. Keep them\nseparate from discovered engine settings.\n\nFor a real-camera brief, direct a credible setting, material condition, motivated light and degree\nof processing. Inspect skin, hands, object joins, reflections and fine texture at full size.\nPhotographic appearance and obeying the requested shot are separate checks; a convincing portrait\ncan still miss its crop or lighting pattern.\n\nFor photography beyond portraits, a named camera brand, or an in-camera processing look, read\n[photography genres and camera cues](references\u002Fphotography-genres.md). Use the genre's actual\nsubject and capture constraints to direct the image; a camera name alone does not establish its\noptics, sensor response, or proprietary color processing.\n\nTo pick a tested style (an art style, an animation school, a film stock, a lens, a director's way of\nshooting, a light), use the [style library](references\u002Fstyle-library.md): each row gives the prompt\nsentence that worked and what to watch for. For style exploration or a nonphotographic brief, read\n[style directions](references\u002Fstyle-directions.md). Choose a specific visual language and describe\nits marks, space, palette, and material. Do not append photographic realism to painting, anime, or\nabstraction unless the user wants a hybrid. Separate a movement, a medium, a genre, and a production\ntechnique rather than treating their names as synonyms.\n\nChoose an offered template when it fits; fill its actual fields with the brief. Choose Looks\ndeliberately rather than filling every row. Avoid contradictory camera or lighting instructions. For\na custom prompt, describe subject and action, framing, setting, light direction and quality, surface\ndetail, and placement of any empty space or exact lettering. State preservation and change\ninstructions explicitly for edits. Use the selected model's official guidance linked from the live\nAPI documentation when its prompting behavior matters.\n\nWhen a prompt has two or more people, in any style, describe each person, not the group: an age,\na face shape, hair, a build and one mark, placed \"from left to right\". Otherwise the faces come back\nas near-clones. A sentence asking for variety does not help. For a family, describe each parent or\ngrandparent the same way, then give each child the traits they take from a named relative (\"a boy\nwith his father's square jaw and wavy black hair\"). Do not write \"they look related\": it dropped a\nperson. See [groups of people](references\u002Fphotography-genres.md#make-each-person-in-a-group-look-different).\n\nFor a coherent set, hold the subject or location, palette, light logic, and treatment steady while\nvarying the requested shot: hero, detail, contextual scene, or alternate crop. A repeated seed does\nnot guarantee the same identity. Once there is a good hero, use it as an Edit reference for\ndependent variations. Generate those only after reviewing that hero. For architecture, preserve the\ndefining materials, openings and layout; a reference image does not establish the geometry hidden\nbehind its visible surfaces.\n\nTell the user before you start when the brief needs something the tested model did not do. On\nQwen-Image-2.1-Turbo (on another backend, treat this list and every Findings table as hypotheses to\ncheck with a concept):\n\n- No transparent background. Generate on pure white and cut it out afterwards.\n- No guaranteed seamless repeat. Keep motifs inside the frame, then tile 2 x 2 and check.\n- No usable animation frames. A sprite sheet is concept art.\n- Edit cannot zoom out or extend the canvas.\n- Rare or technical Chinese, Japanese or Korean characters come out malformed at any size.\n- Technical diagrams look right but are not accurate. Do not use them as reference material.\n- Chart bars are not proportional to their values. Draw a data chart with code.\n- Grid counts drift. A sheet of 8 icons can come back with 9, and tiles do not connect.\n- Quoted names can lose their letter order (\"Sheryl\" came back \"Sherly\"). Check names letter by\n  letter.\n- Small Japanese, Korean or Arabic lines come out with wrong characters, even when the headline\n  above them is exact. Set small non-Latin copy in a layout tool.\n\nUse a small concept when the brief and budget allow it. Compare one meaningful variable at a time\nwhen testing a style. Keep known settings and seed fixed for a controlled comparison. A smaller\nconcept helps composition decisions but cannot prove fine text or final-resolution detail. If the\nuser declines a concept, tell them which styles in the brief the style library marks partial or does\nnot list, and what to watch for in the final. Changing dimensions can change composition; do not\ndescribe a new render as a pixel upscale.\n\nBatch independent shots with one backend when useful. Do not put an edit that needs an earlier\noutput into the same batch: the reference bytes do not exist yet. Save requests and operation IDs\nbefore continuing, and download completed images promptly, including partial successes.\n\nDo not switch to the full model only to raise the quality. Six finished Turbo prompts, made again\non full Qwen-Image-2.1 at its own 40 steps, took about five times as long. Two came back richer\n(finer stained glass, more dragons), three about the same, and one worse (the jellyfish turned into\nflat saucers). Offer the full model as another try at a select, and compare both.\n\n## Inspect, select, and refine\n\nInspect saved images, not just API status. A contact sheet helps choose a direction; inspect selects\nindividually at full size for the details that matter. Check against the brief:\n\n- Subject identity, product shape, proportions, label spelling, and material finish.\n- Hands and contact points, plausible shadows\u002Freflections, duplicated or merged objects.\n- For complex scenes, exact inventories, containment and object-to-object relationships; separate\n  physical objects from their reflections and occluded parts.\n- Framing, edge cutoffs, intended aspect ratio, useful copy space, and readable exact text.\n- Requested camera height and viewpoint, relative object scale, focus target, blur direction, and\n  the light\u002Fshadow arrangement; do not infer adherence from a lens name in metadata.\n- Consistency across the set without accidentally repeating every composition.\n\nAdapt these checks to the intended language: flattened space or distorted anatomy may be deliberate.\nJudge whether the image follows the chosen direction before calling those defects. For style\ncomparisons, assess the rendering method separately from subject retention.\n\nKeep the chosen original. Make a targeted edit with a concrete change and explicit preservation\ninstructions; compare it with its source. A language instruction to change a region is not a mask.\nWhere the mode offers region marking, a mark places a change precisely (see\n[editing](references\u002Fediting.md)), but it does not protect the rest of the frame either. Do not\npromise that pixels outside any edit stay unchanged.\n\nStop when the requested set meets the brief or its budget is exhausted. Report remaining defects\ninstead of silently launching more attempts. Preserve successful items if a later item fails.\n\n## Deliver\n\nSave selects with descriptive filenames and report their paths, dimensions, and useful selection\nnotes. For a multi-image shoot, keep a compact record linking each select to its request,\noperation\u002Fitem ID, seed, reference files, and edits; retain the original references. A simple\nsingle-image task only needs enough provenance to reproduce or revise it. Separate requested finals\nfrom concepts and rejected attempts. Mention unresolved fidelity or text issues. Use local links or\nthe user's requested destination; publishing is a separate action. Run the finishing steps the end\nuse needs (cutout, trace, resize, upscale, alt text) from [delivery](references\u002Fdelivery.md), and\nsay which ones you ran and which the user still has to.\n\n## Words used here\n\n- **Brief**: what the user wants, in their words: subject, use, style.\n- **Concept**: a small, fast image that shows one possible direction before the final.\n- **Select**: an image the user chose to keep.\n- **Hero image**: the main image of a page, banner or set.\n- **Style library**: the [table of tested styles](references\u002Fstyle-library.md), one prompt\n  sentence each.\n- **Looks** and **Rewrite**: the studio's own features. Looks adds preset style text to the\n  prompt; Rewrite expands a short request into a full prompt.\n- **Accuracy level**: how true the image must be: mood only, roughly right, or must be exact.\n- **Safe area**: the part of a banner or post that every device shows uncropped.\n- **Dieline**: the flat cutting template of a package, drawn by a designer, not by the model.\n","# Studio Image\n\nTurn a brief into still images that work together, then inspect and refine the actual images.\nOperate the user's Open Imagine server through its public API. Keep this workflow usable with\ndifferent Image models; discover their capabilities rather than assuming a particular checkpoint.\nRequires a reachable server and tools for HTTP, saving images, and visual inspection.\n\n## Choose the path\n\nIf the request is short or vague (\"a nice photo of my dog\"), or the user says they do not know where\nto start, follow the [brief-to-final flow](references\u002Fbrief-to-final.md). It asks at most three\nplain questions, shows four concepts from the [style library](references\u002Fstyle-library.md), and\nmakes the final from the concept the user picks. A user with a full brief (subject, use, style)\ncontinues below.\n\n## Establish the shoot\n\nUse the user's server URL; otherwise try `http:\u002F\u002F127.0.0.1:8765`. Read [the API operating\nnotes](references\u002Fstudio-api.md) before submitting work. If the server is unavailable, report the\nconnection problem; starting or installing a server is a separate task unless already authorized.\nOnce the backend is known, read [model notes](references\u002Fmodel-notes.md) for measured behavior on\nthat checkpoint. An unlisted model is untested; use its live capabilities and its own prompt\nwording.\n\nExtract the subject, intended use, deliverable count, aspect ratios, and visual direction from the\nbrief. Identify what must remain exact: a face, product silhouette, label, material, garment, or\nsupplied composition. Use sensible defaults for missing creative details; ask only when an unknown\nchanges the requested result materially. Respect the user's generation\u002Ftime budget. A request for\none finished image does not imply permission for an extensive concept round (the brief-to-final\nflow's four small concepts are its normal start, not an extra batch).\n\nGive each reference a role: subject identity, product geometry, wardrobe, lighting, pose, or layout.\nRefer to images in their submitted order using the selected model's supported wording. Distinguish a\nreference's required content from incidental background details. If a supplied product or person\nmust remain recognizable, prefer reference-based Edit over reinventing it from text. Do not claim\nexact brand or identity fidelity without inspecting the result.\n\nBefore you write the prompt, check the request against [boundaries](references\u002Fboundaries.md) if it\ninvolves a real person, a document, someone else's work, a brand or a mark, health, or news. Check\nit before Rewrite, because Rewrite carries the request into a longer prompt.\n\nAsk for the end use (screen, print, sticker, platform, vector, cutout) when the brief does not imply\nit, and otherwise state the end use you assume, because it sets the background, the margins and the\nsize; [delivery](references\u002Fdelivery.md) has the recipes. Split the job: the model makes the\npicture, and a layout tool or script adds what must be exact (long copy, exact colors, QR codes,\ngrids, counts, cut lines, the final pixel size, transparency).\n\nPlace each request on one accuracy level, and tell the user which:\n\n- **Mood only.** Mood, style and composition. The model is fine.\n- **Roughly right.** History, botany, products, architecture: it looks right and needs a human check\n  before anyone relies on it.\n- **Must be exact.** Data charts, maps with real geography, anatomy for teaching, floor plans, math\n  and chemistry figures, sewing patterns. Draw these with code or CAD, not this model.\n\n## Direct and generate\n\nFind the brief's kind of image in [use cases](references\u002Fuse-cases.md). It names the file to read, a\ndefault ratio, and whether Rewrite helped on the tested model. Before you write a custom prompt or\nuse Rewrite, read [prompt craft](references\u002Fprompt-craft.md). For an image that carries words or a\nlayout, read [design and text](references\u002Fdesign-and-text.md). For stickers, icons, characters,\ncomics, sprites or other drawn assets, read [illustration and\nassets](references\u002Fillustration-and-assets.md). For any Edit, read [editing](references\u002Fediting.md).\n\nFor composition, viewpoint, framing, focus, camera effects, lighting, or subject staging, read [shot\ndesign](references\u002Fshot-design.md). Treat these as independent decisions from style. Choose the\nvisible outcome first, then use compatible camera shorthand to support it. Camera numbers in prompt\ntext are creative cues, not calibrated API controls or proof of an actual capture setup. Keep them\nseparate from discovered engine settings.\n\nFor a real-camera brief, direct a credible setting, material condition, motivated light and degree\nof processing. Inspect skin, hands, object joins, reflections and fine texture at full size.\nPhotographic appearance and obeying the requested shot are separate checks; a convincing portrait\ncan still miss its crop or lighting pattern.\n\nFor photography beyond portraits, a named camera brand, or an in-camera processing look, read\n[photography genres and camera cues](references\u002Fphotography-genres.md). Use the genre's actual\nsubject and capture constraints to direct the image; a camera name alone does not establish its\noptics, sensor response, or proprietary color processing.\n\nTo pick a tested style (an art style, an animation school, a film stock, a lens, a director's way of\nshooting, a light), use the [style library](references\u002Fstyle-library.md): each row gives the prompt\nsentence that worked and what to watch for. For style exploration or a nonphotographic brief, read\n[style directions](references\u002Fstyle-directions.md). Choose a specific visual language and describe\nits marks, space, palette, and material. Do not append photographic realism to painting, anime, or\nabstraction unless the user wants a hybrid. Separate a movement, a medium, a genre, and a production\ntechnique rather than treating their names as synonyms.\n\nChoose an offered template when it fits; fill its actual fields with the brief. Choose Looks\ndeliberately rather than filling every row. Avoid contradictory camera or lighting instructions. For\na custom prompt, describe subject and action, framing, setting, light direction and quality, surface\ndetail, and placement of any empty space or exact lettering. State preservation and change\ninstructions explicitly for edits. Use the selected model's official guidance linked from the live\nAPI documentation when its prompting behavior matters.\n\nWhen a prompt has two or more people, in any style, describe each person, not the group: an age,\na face shape, hair, a build and one mark, placed \"from left to right\". Otherwise the faces come back\nas near-clones. A sentence asking for variety does not help. For a family, describe each parent or\ngrandparent the same way, then give each child the traits they take from a named relative (\"a boy\nwith his father's square jaw and wavy black hair\"). Do not write \"they look related\": it dropped a\nperson. See [groups of people](references\u002Fphotography-genres.md#make-each-person-in-a-group-look-different).\n\nFor a coherent set, hold the subject or location, palette, light logic, and treatment steady while\nvarying the requested shot: hero, detail, contextual scene, or alternate crop. A repeated seed does\nnot guarantee the same identity. Once there is a good hero, use it as an Edit reference for\ndependent variations. Generate those only after reviewing that hero. For architecture, preserve the\ndefining materials, openings and layout; a reference image does not establish the geometry hidden\nbehind its visible surfaces.\n\nTell the user before you start when the brief needs something the tested model did not do. On\nQwen-Image-2.1-Turbo (on another backend, treat this list and every Findings table as hypotheses to\ncheck with a concept):\n\n- No transparent background. Generate on pure white and cut it out afterwards.\n- No guaranteed seamless repeat. Keep motifs inside the frame, then tile 2 x 2 and check.\n- No usable animation frames. A sprite sheet is concept art.\n- Edit cannot zoom out or extend the canvas.\n- Rare or technical Chinese, Japanese or Korean characters come out malformed at any size.\n- Technical diagrams look right but are not accurate. Do not use them as reference material.\n- Chart bars are not proportional to their values. Draw a data chart with code.\n- Grid counts drift. A sheet of 8 icons can come back with 9, and tiles do not connect.\n- Quoted names can lose their letter order (\"Sheryl\" came back \"Sherly\"). Check names letter by\n  letter.\n- Small Japanese, Korean or Arabic lines come out with wrong characters, even when the headline\n  above them is exact. Set small non-Latin copy in a layout tool.\n\nUse a small concept when the brief and budget allow it. Compare one meaningful variable at a time\nwhen testing a style. Keep known settings and seed fixed for a controlled comparison. A smaller\nconcept helps composition decisions but cannot prove fine text or final-resolution detail. If the\nuser declines a concept, tell them which styles in the brief the style library marks partial or does\nnot list, and what to watch for in the final. Changing dimensions can change composition; do not\ndescribe a new render as a pixel upscale.\n\nBatch independent shots with one backend when useful. Do not put an edit that needs an earlier\noutput into the same batch: the reference bytes do not exist yet. Save requests and operation IDs\nbefore continuing, and download completed images promptly, including partial successes.\n\nDo not switch to the full model only to raise the quality. Six finished Turbo prompts, made again\non full Qwen-Image-2.1 at its own 40 steps, took about five times as long. Two came back richer\n(finer stained glass, more dragons), three about the same, and one worse (the jellyfish turned into\nflat saucers). Offer the full model as another try at a select, and compare both.\n\n## Inspect, select, and refine\n\nInspect saved images, not just API status. A contact sheet helps choose a direction; inspect selects\nindividually at full size for the details that matter. Check against the brief:\n\n- Subject identity, product shape, proportions, label spelling, and material finish.\n- Hands and contact points, plausible shadows\u002Freflections, duplicated or merged objects.\n- For complex scenes, exact inventories, containment and object-to-object relationships; separate\n  physical objects from their reflections and occluded parts.\n- Framing, edge cutoffs, intended aspect ratio, useful copy space, and readable exact text.\n- Requested camera height and viewpoint, relative object scale, focus target, blur direction, and\n  the light\u002Fshadow arrangement; do not infer adherence from a lens name in metadata.\n- Consistency across the set without accidentally repeating every composition.\n\nAdapt these checks to the intended language: flattened space or distorted anatomy may be deliberate.\nJudge whether the image follows the chosen direction before calling those defects. For style\ncomparisons, assess the rendering method separately from subject retention.\n\nKeep the chosen original. Make a targeted edit with a concrete change and explicit preservation\ninstructions; compare it with its source. A language instruction to change a region is not a mask.\nWhere the mode offers region marking, a mark places a change precisely (see\n[editing](references\u002Fediting.md)), but it does not protect the rest of the frame either. Do not\npromise that pixels outside any edit stay unchanged.\n\nStop when the requested set meets the brief or its budget is exhausted. Report remaining defects\ninstead of silently launching more attempts. Preserve successful items if a later item fails.\n\n## Deliver\n\nSave selects with descriptive filenames and report their paths, dimensions, and useful selection\nnotes. For a multi-image shoot, keep a compact record linking each select to its request,\noperation\u002Fitem ID, seed, reference files, and edits; retain the original references. A simple\nsingle-image task only needs enough provenance to reproduce or revise it. Separate requested finals\nfrom concepts and rejected attempts. Mention unresolved fidelity or text issues. Use local links or\nthe user's requested destination; publishing is a separate action. Run the finishing steps the end\nuse needs (cutout, trace, resize, upscale, alt text) from [delivery](references\u002Fdelivery.md), and\nsay which ones you ran and which the user still has to.\n\n## Words used here\n\n- **Brief**: what the user wants, in their words: subject, use, style.\n- **Concept**: a small, fast image that shows one possible direction before the final.\n- **Select**: an image the user chose to keep.\n- **Hero image**: the main image of a page, banner or set.\n- **Style library**: the [table of tested styles](references\u002Fstyle-library.md), one prompt\n  sentence each.\n- **Looks** and **Rewrite**: the studio's own features. Looks adds preset style text to the\n  prompt; Rewrite expands a short request into a full prompt.\n- **Accuracy level**: how true the image must be: mood only, roughly right, or must be exact.\n- **Safe area**: the part of a banner or post that every device shows uncropped.\n- **Dieline**: the flat cutting template of a package, drawn by a designer, not by the model.\n",{"name":5,"description":7},"main","5b43f087128f7e63a681adf39f7fbf412fc93987","plugins\u002Fopen-imagine\u002Fskills\u002Fstudio-image",1,"https:\u002F\u002Fgithub.com\u002FCJHwong\u002Fopen-imagine","https:\u002F\u002Fcjhwong.github.io\u002Fopen-imagine-gallery\u002F","https:\u002F\u002Fgithub.com\u002FCJHwong","2026-11-14 02:31:40",{"source":8,"skillName":5,"commands":33,"prompt":43,"zipAvailable":44},[34,37,40],{"tool":35,"command":36},"npx","npx skills add CJHwong\u002Fopen-imagine --skill studio-image",{"tool":38,"command":39},"bunx","bunx skills add CJHwong\u002Fopen-imagine --skill studio-image",{"tool":41,"command":42},"pnpm","pnpm dlx skills add CJHwong\u002Fopen-imagine --skill studio-image","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FCJHwong\u002Fopen-imagine\u002Fblob\u002Fmain\u002Fplugins\u002Fopen-imagine\u002Fskills\u002Fstudio-image\u002FSKILL.md",true,[46,52,56,60,64,68,72,76,80,84,88,92,96,100,104],{"path":47,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":50,"sourceHtmlUrl":51},"agents\u002Fopenai.yaml","REFERENCE","参考文档",127,"https:\u002F\u002Fgithub.com\u002FCJHwong\u002Fopen-imagine\u002Fblob\u002Fmain\u002Fplugins\u002Fopen-imagine\u002Fskills\u002Fstudio-image\u002Fagents\u002Fopenai.yaml",{"path":53,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":54,"sourceHtmlUrl":55},"references\u002Fboundaries.md",4801,"https:\u002F\u002Fgithub.com\u002FCJHwong\u002Fopen-imagine\u002Fblob\u002Fmain\u002Fplugins\u002Fopen-imagine\u002Fskills\u002Fstudio-image\u002Freferences\u002Fboundaries.md",{"path":57,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":58,"sourceHtmlUrl":59},"references\u002Fbrief-to-final.md",7960,"https:\u002F\u002Fgithub.com\u002FCJHwong\u002Fopen-imagine\u002Fblob\u002Fmain\u002Fplugins\u002Fopen-imagine\u002Fskills\u002Fstudio-image\u002Freferences\u002Fbrief-to-final.md",{"path":61,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":62,"sourceHtmlUrl":63},"references\u002Fdelivery.md",7120,"https:\u002F\u002Fgithub.com\u002FCJHwong\u002Fopen-imagine\u002Fblob\u002Fmain\u002Fplugins\u002Fopen-imagine\u002Fskills\u002Fstudio-image\u002Freferences\u002Fdelivery.md",{"path":65,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":66,"sourceHtmlUrl":67},"references\u002Fdesign-and-text.md",10111,"https:\u002F\u002Fgithub.com\u002FCJHwong\u002Fopen-imagine\u002Fblob\u002Fmain\u002Fplugins\u002Fopen-imagine\u002Fskills\u002Fstudio-image\u002Freferences\u002Fdesign-and-text.md",{"path":69,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":70,"sourceHtmlUrl":71},"references\u002Fediting.md",6758,"https:\u002F\u002Fgithub.com\u002FCJHwong\u002Fopen-imagine\u002Fblob\u002Fmain\u002Fplugins\u002Fopen-imagine\u002Fskills\u002Fstudio-image\u002Freferences\u002Fediting.md",{"path":73,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":74,"sourceHtmlUrl":75},"references\u002Fillustration-and-assets.md",8621,"https:\u002F\u002Fgithub.com\u002FCJHwong\u002Fopen-imagine\u002Fblob\u002Fmain\u002Fplugins\u002Fopen-imagine\u002Fskills\u002Fstudio-image\u002Freferences\u002Fillustration-and-assets.md",{"path":77,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":78,"sourceHtmlUrl":79},"references\u002Fmodel-notes.md",3344,"https:\u002F\u002Fgithub.com\u002FCJHwong\u002Fopen-imagine\u002Fblob\u002Fmain\u002Fplugins\u002Fopen-imagine\u002Fskills\u002Fstudio-image\u002Freferences\u002Fmodel-notes.md",{"path":81,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":82,"sourceHtmlUrl":83},"references\u002Fphotography-genres.md",24217,"https:\u002F\u002Fgithub.com\u002FCJHwong\u002Fopen-imagine\u002Fblob\u002Fmain\u002Fplugins\u002Fopen-imagine\u002Fskills\u002Fstudio-image\u002Freferences\u002Fphotography-genres.md",{"path":85,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":86,"sourceHtmlUrl":87},"references\u002Fprompt-craft.md",7222,"https:\u002F\u002Fgithub.com\u002FCJHwong\u002Fopen-imagine\u002Fblob\u002Fmain\u002Fplugins\u002Fopen-imagine\u002Fskills\u002Fstudio-image\u002Freferences\u002Fprompt-craft.md",{"path":89,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":90,"sourceHtmlUrl":91},"references\u002Fshot-design.md",23054,"https:\u002F\u002Fgithub.com\u002FCJHwong\u002Fopen-imagine\u002Fblob\u002Fmain\u002Fplugins\u002Fopen-imagine\u002Fskills\u002Fstudio-image\u002Freferences\u002Fshot-design.md",{"path":93,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":94,"sourceHtmlUrl":95},"references\u002Fstudio-api.md",9495,"https:\u002F\u002Fgithub.com\u002FCJHwong\u002Fopen-imagine\u002Fblob\u002Fmain\u002Fplugins\u002Fopen-imagine\u002Fskills\u002Fstudio-image\u002Freferences\u002Fstudio-api.md",{"path":97,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":98,"sourceHtmlUrl":99},"references\u002Fstyle-directions.md",21835,"https:\u002F\u002Fgithub.com\u002FCJHwong\u002Fopen-imagine\u002Fblob\u002Fmain\u002Fplugins\u002Fopen-imagine\u002Fskills\u002Fstudio-image\u002Freferences\u002Fstyle-directions.md",{"path":101,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":102,"sourceHtmlUrl":103},"references\u002Fstyle-library.md",55986,"https:\u002F\u002Fgithub.com\u002FCJHwong\u002Fopen-imagine\u002Fblob\u002Fmain\u002Fplugins\u002Fopen-imagine\u002Fskills\u002Fstudio-image\u002Freferences\u002Fstyle-library.md",{"path":105,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":106,"sourceHtmlUrl":107},"references\u002Fuse-cases.md",19296,"https:\u002F\u002Fgithub.com\u002FCJHwong\u002Fopen-imagine\u002Fblob\u002Fmain\u002Fplugins\u002Fopen-imagine\u002Fskills\u002Fstudio-image\u002Freferences\u002Fuse-cases.md",[109],{"skillNo":110,"name":111,"nameZh":6,"description":112,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":113,"categories":114,"keywords":115,"publishedAt":19},"01M799TNE29MYGY20WJERENKV9","studio-music","Create songs, instrumentals, covers and scores with Open Imagine's Music API. Write and adapt lyrics, choose planning and an ABC score, transcribe a recording, edit a score while preserving the melody, and compare takes by listening. Use for song requests, first songs, lyric writing, instrumental cues, covers, reharmonization, score editing, shuffled song sets and listening comparisons, and when the user does not know where to start and wants to be guided; also for 歌曲生成, 純音樂, 翻唱, 轉譜, 改譜 and 歌詞改寫.","https:\u002F\u002Fgithub.com\u002FCJHwong\u002Fopen-imagine\u002Fblob\u002Fmain\u002Fplugins\u002Fopen-imagine\u002Fskills\u002Fstudio-music\u002FSKILL.md",[],[]]