[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6JQDBMFRGP4X7MRZG0JYNHT":3,"skill-same:01M6JQDBMFRGP4X7MRZG0JYNHT":93},{"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":23,"publishedAt":41,"linkOnly":42,"markdownContent":43,"markdownBody":44,"frontMatter":45,"refName":46,"commitSha":47,"skillDirectory":48,"forksCount":49,"repositoryHtmlUrl":50,"repositoryHomepageUrl":51,"ownerHtmlUrl":52,"ownerBlogUrl":53,"ownerDescription":6,"updatedAt":54,"install":55,"resources":68},"01M6JQDBMFRGP4X7MRZG0JYNHT","sglang_diffusion_ako4all_kernel",null,"Use when optimizing an existing SGLang diffusion kernel with AKO4ALL, including AKO4ALL repo hygiene, custom microbench setup, ncu-guided iteration, and end-to-end denoise validation. Also use when a sibling AKO4ALL repo must be cloned or refreshed before starting kernel tuning work.","LongLeo287\u002FOmniClaw","LongLeo287","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F141803367?v=4","01M6JK75C62VCNZKR5BRZ386YV","LongLeo","01M6JK75GQ78TSP1ZPXNKK3MCH",9,"MIT","https:\u002F\u002Fgithub.com\u002FLongLeo287\u002FOmniClaw\u002Fblob\u002Fmain\u002Fecosystem\u002Fskills\u002Fsglang_diffusion_ako4all_kernel\u002FSKILL.md",[18],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",[24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40],"sglang diffusion ako4all kernel","SGLang","diffusion","kernel","iteration","custom","Existing","with","denoise","hygiene","end","use","repo","work","tuning","validation","setup","2026-11-10 10:12:30",false,"---\nname: sglang_diffusion_ako4all_kernel\n\ndescription: Use when optimizing an existing SGLang diffusion kernel with AKO4ALL, including AKO4ALL repo hygiene, custom microbench setup, ncu-guided iteration, and end-to-end denoise validation. Also use when a sibling AKO4ALL repo must be cloned or refreshed before starting kernel tuning work.\n---\n\n# SGLang Diffusion AKO4ALL Kernel\n\nUse this skill to run the full AKO4ALL-based optimization loop for an existing SGLang diffusion kernel.\nIt is the default implementation path once the benchmark\u002Fprofile skill has already shown that a hotspot is real and not covered by an existing fast path. This workflow bootstraps a custom AKO harness, benchmarks and profiles the kernel, iterates with `ncu`, ports the best version back to `sglang`, then validates with targeted tests and model-level denoise runs.\n\nThis skill assumes a sibling repo layout like:\n\n```text\n\u003Cbase-dir>\u002F\n├── sglang\u002F\n└── AKO4ALL\u002F\n```\n\nIf `AKO4ALL\u002F` is missing under the current base directory, clone it first.\n\n## Use This Skill When\n\n- tuning an existing diffusion Triton, CUDA JIT, CuTeDSL, or runtime-integrated kernel in `sglang`\n- `sglang-diffusion-benchmark-profile` has already ruled out an existing in-repo fast path or overlap family\n- creating a custom AKO4ALL harness for a real diffusion kernel instead of using the default benchmark tasks\n- validating that a kernel-level win transfers to Qwen, FLUX, Wan, Hunyuan, MOVA, or other diffusion denoise latency\n- preparing PR artifacts such as microbench tables, `ncu` before\u002Fafter data, and proof image outputs\n\nDo not start here when the bottleneck has not been proven yet.\nFirst use [..\u002Fsglang-diffusion-benchmark-profile\u002FSKILL.md](..\u002Fsglang-diffusion-benchmark-profile\u002FSKILL.md) to:\n- measure the real denoise regression\n- collect the perf dump baseline\n- capture one representative `torch.profiler` trace\n- rule out existing merged fast paths\n\nIf a future specialized optimization skill matches the kernel family better than AKO4ALL, hand off there instead. The diagnosis contract stays the same.\n\n## Mandatory AKO4ALL Preflight\n\nBefore any AKO work:\n\n1. Run `scripts\u002Fensure_ako4all_clean.sh [base-dir]`.\n2. If `\u003Cbase-dir>\u002FAKO4ALL` does not exist, the script clones it.\n3. Do not continue unless `AKO4ALL` is:\n   - on the upstream default branch, usually `main`\n   - fully clean with no tracked or untracked local changes\n   - exactly synced to `upstream\u002F\u003Cdefault-branch>`\n4. If the script reports local commits, divergence, or a dirty worktree, stop and clean or re-clone the repo before continuing.\n\nThe script creates an `upstream` remote automatically when missing.\nBy default it uses the existing `origin` URL, or `AKO4ALL_URL` if you need to override the clone source.\n\n## Workflow\n\n### 1. Scope the Kernel\n\n- Identify the exact kernel entry point and runtime call sites in `sglang`.\n- Record the target shapes, dtypes, model families, and whether the kernel is on a hot path.\n- Reuse existing unit tests and benchmark entry points when they already exist.\n\n### 2. Bootstrap the AKO Harness\n\nInside the clean `AKO4ALL` repo:\n\n- read `TASK.md` and `HINTS.md`\n- create a custom harness instead of relying on the stock benchmark tasks\n- mirror the real SGLang kernel into:\n  - `input\u002Freference.py`\n  - `input\u002F\u003Ckernel>.py`\n  - `solution\u002F\u003Ckernel>.py`\n  - `bench\u002Fbench_\u003Ckernel>.py`\n- keep a short context note in `context\u002F` when the kernel has model-specific shape assumptions or perf conclusions\n\nThe custom benchmark should:\n\n- cover representative diffusion shapes\n- check correctness against the reference kernel\n- report aggregate runtime plus per-shape results when useful\n\n### 3. Establish the Baseline\n\n- run the AKO custom microbench before changing the kernel\n- capture one representative `ncu` baseline on the hottest meaningful shape\n- note whether the bottleneck looks like registers, occupancy, instruction count, launch config, or memory latency\n\n### 4. Iterate in AKO4ALL\n\n- change one idea at a time\n- rerun the microbench after every change\n- update `ITERATIONS.md` with hypothesis, result, and next step\n- prefer simple, explainable wins over clever rewrites that do not transfer\n\nAfter 3 consecutive no-improvement or regression iterations:\n\n- rerun `ncu`\n- re-read `ITERATIONS.md`\n- change direction instead of continuing blind sweeps\n\n### 5. Port the Best Version Back to SGLang\n\n- apply the best candidate to the real `sglang` kernel file\n- run import or syntax checks and targeted tests first\n- keep the AKO `solution\u002F` version aligned with the main-tree version you actually want to keep\n\n### 6. Validate on Real Models\n\n- use the benchmark\u002Fprofile skill for denoise perf dumps and before\u002Fafter comparison\n- prefer exact local snapshot validation when testing local edits on a GPU box\n- run targeted kernel tests first\n- run model-level denoise benchmarks with perf dumps\n- compare baseline vs optimized runs with `compare_perf.py`\n- if the PR needs proof that generation still works, save one real model output image\n\n### 7. Prepare PR Artifacts\n\nAt minimum, keep:\n\n- one microbench table\n- one denoise-stage table\n- one end-to-end table\n- one `ncu` before\u002Fafter pair on the most representative kernel shape\n- one generated image when the kernel affects production inference\n\nSee [references\u002Fako-loop.md](references\u002Fako-loop.md) for the checklist and common stop rules.\n\n## Operating Rules\n\n- Treat AKO4ALL repo hygiene as a gate, not a suggestion.\n- Prefer exact local snapshot validation over hand-wavy “remote tree is close enough”.\n- Keep model-level validation honest: if microbench improves but denoise does not, do not keep the AKO-only variant in the main code path.\n- When writing conclusions, explain the win in terms of measurable causes such as lower registers per thread, higher occupancy, fewer executed instructions, or better scheduler eligibility.\n","# SGLang Diffusion AKO4ALL Kernel\n\nUse this skill to run the full AKO4ALL-based optimization loop for an existing SGLang diffusion kernel.\nIt is the default implementation path once the benchmark\u002Fprofile skill has already shown that a hotspot is real and not covered by an existing fast path. This workflow bootstraps a custom AKO harness, benchmarks and profiles the kernel, iterates with `ncu`, ports the best version back to `sglang`, then validates with targeted tests and model-level denoise runs.\n\nThis skill assumes a sibling repo layout like:\n\n```text\n\u003Cbase-dir>\u002F\n├── sglang\u002F\n└── AKO4ALL\u002F\n```\n\nIf `AKO4ALL\u002F` is missing under the current base directory, clone it first.\n\n## Use This Skill When\n\n- tuning an existing diffusion Triton, CUDA JIT, CuTeDSL, or runtime-integrated kernel in `sglang`\n- `sglang-diffusion-benchmark-profile` has already ruled out an existing in-repo fast path or overlap family\n- creating a custom AKO4ALL harness for a real diffusion kernel instead of using the default benchmark tasks\n- validating that a kernel-level win transfers to Qwen, FLUX, Wan, Hunyuan, MOVA, or other diffusion denoise latency\n- preparing PR artifacts such as microbench tables, `ncu` before\u002Fafter data, and proof image outputs\n\nDo not start here when the bottleneck has not been proven yet.\nFirst use [..\u002Fsglang-diffusion-benchmark-profile\u002FSKILL.md](..\u002Fsglang-diffusion-benchmark-profile\u002FSKILL.md) to:\n- measure the real denoise regression\n- collect the perf dump baseline\n- capture one representative `torch.profiler` trace\n- rule out existing merged fast paths\n\nIf a future specialized optimization skill matches the kernel family better than AKO4ALL, hand off there instead. The diagnosis contract stays the same.\n\n## Mandatory AKO4ALL Preflight\n\nBefore any AKO work:\n\n1. Run `scripts\u002Fensure_ako4all_clean.sh [base-dir]`.\n2. If `\u003Cbase-dir>\u002FAKO4ALL` does not exist, the script clones it.\n3. Do not continue unless `AKO4ALL` is:\n   - on the upstream default branch, usually `main`\n   - fully clean with no tracked or untracked local changes\n   - exactly synced to `upstream\u002F\u003Cdefault-branch>`\n4. If the script reports local commits, divergence, or a dirty worktree, stop and clean or re-clone the repo before continuing.\n\nThe script creates an `upstream` remote automatically when missing.\nBy default it uses the existing `origin` URL, or `AKO4ALL_URL` if you need to override the clone source.\n\n## Workflow\n\n### 1. Scope the Kernel\n\n- Identify the exact kernel entry point and runtime call sites in `sglang`.\n- Record the target shapes, dtypes, model families, and whether the kernel is on a hot path.\n- Reuse existing unit tests and benchmark entry points when they already exist.\n\n### 2. Bootstrap the AKO Harness\n\nInside the clean `AKO4ALL` repo:\n\n- read `TASK.md` and `HINTS.md`\n- create a custom harness instead of relying on the stock benchmark tasks\n- mirror the real SGLang kernel into:\n  - `input\u002Freference.py`\n  - `input\u002F\u003Ckernel>.py`\n  - `solution\u002F\u003Ckernel>.py`\n  - `bench\u002Fbench_\u003Ckernel>.py`\n- keep a short context note in `context\u002F` when the kernel has model-specific shape assumptions or perf conclusions\n\nThe custom benchmark should:\n\n- cover representative diffusion shapes\n- check correctness against the reference kernel\n- report aggregate runtime plus per-shape results when useful\n\n### 3. Establish the Baseline\n\n- run the AKO custom microbench before changing the kernel\n- capture one representative `ncu` baseline on the hottest meaningful shape\n- note whether the bottleneck looks like registers, occupancy, instruction count, launch config, or memory latency\n\n### 4. Iterate in AKO4ALL\n\n- change one idea at a time\n- rerun the microbench after every change\n- update `ITERATIONS.md` with hypothesis, result, and next step\n- prefer simple, explainable wins over clever rewrites that do not transfer\n\nAfter 3 consecutive no-improvement or regression iterations:\n\n- rerun `ncu`\n- re-read `ITERATIONS.md`\n- change direction instead of continuing blind sweeps\n\n### 5. Port the Best Version Back to SGLang\n\n- apply the best candidate to the real `sglang` kernel file\n- run import or syntax checks and targeted tests first\n- keep the AKO `solution\u002F` version aligned with the main-tree version you actually want to keep\n\n### 6. Validate on Real Models\n\n- use the benchmark\u002Fprofile skill for denoise perf dumps and before\u002Fafter comparison\n- prefer exact local snapshot validation when testing local edits on a GPU box\n- run targeted kernel tests first\n- run model-level denoise benchmarks with perf dumps\n- compare baseline vs optimized runs with `compare_perf.py`\n- if the PR needs proof that generation still works, save one real model output image\n\n### 7. Prepare PR Artifacts\n\nAt minimum, keep:\n\n- one microbench table\n- one denoise-stage table\n- one end-to-end table\n- one `ncu` before\u002Fafter pair on the most representative kernel shape\n- one generated image when the kernel affects production inference\n\nSee [references\u002Fako-loop.md](references\u002Fako-loop.md) for the checklist and common stop rules.\n\n## Operating Rules\n\n- Treat AKO4ALL repo hygiene as a gate, not a suggestion.\n- Prefer exact local snapshot validation over hand-wavy “remote tree is close enough”.\n- Keep model-level validation honest: if microbench improves but denoise does not, do not keep the AKO-only variant in the main code path.\n- When writing conclusions, explain the win in terms of measurable causes such as lower registers per thread, higher occupancy, fewer executed instructions, or better scheduler eligibility.\n",{"name":5,"description":7},"main","c173cdffe08e788a99b91928a831b476b8434838","ecosystem\u002Fskills\u002Fsglang_diffusion_ako4all_kernel",0,"https:\u002F\u002Fgithub.com\u002FLongLeo287\u002FOmniClaw","https:\u002F\u002Fportfolio-longleo.vercel.app\u002Flabs\u002Fomniclaw.html","https:\u002F\u002Fgithub.com\u002FLongLeo287","https:\u002F\u002Fwww.facebook.com\u002FResourceRookie2023","2026-09-30 00:58:05",{"source":8,"skillName":5,"commands":56,"prompt":66,"zipAvailable":67},[57,60,63],{"tool":58,"command":59},"npx","npx skills add LongLeo287\u002FOmniClaw --skill sglang_diffusion_ako4all_kernel",{"tool":61,"command":62},"bunx","bunx skills add LongLeo287\u002FOmniClaw --skill sglang_diffusion_ako4all_kernel",{"tool":64,"command":65},"pnpm","pnpm dlx skills add LongLeo287\u002FOmniClaw --skill sglang_diffusion_ako4all_kernel","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FLongLeo287\u002FOmniClaw\u002Fblob\u002Fmain\u002Fecosystem\u002Fskills\u002Fsglang_diffusion_ako4all_kernel\u002FSKILL.md",true,[69,75,79,83,87],{"path":70,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":73,"sourceHtmlUrl":74},"_DIR_IDENTITY.md","REFERENCE","参考文档",184,"https:\u002F\u002Fgithub.com\u002FLongLeo287\u002FOmniClaw\u002Fblob\u002Fmain\u002Fecosystem\u002Fskills\u002Fsglang_diffusion_ako4all_kernel\u002F_DIR_IDENTITY.md",{"path":76,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":77,"sourceHtmlUrl":78},"README.md",497,"https:\u002F\u002Fgithub.com\u002FLongLeo287\u002FOmniClaw\u002Fblob\u002Fmain\u002Fecosystem\u002Fskills\u002Fsglang_diffusion_ako4all_kernel\u002FREADME.md",{"path":80,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":81,"sourceHtmlUrl":82},"references\u002Fako_loop.md",1433,"https:\u002F\u002Fgithub.com\u002FLongLeo287\u002FOmniClaw\u002Fblob\u002Fmain\u002Fecosystem\u002Fskills\u002Fsglang_diffusion_ako4all_kernel\u002Freferences\u002Fako_loop.md",{"path":84,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":85,"sourceHtmlUrl":86},"schema.json",495,"https:\u002F\u002Fgithub.com\u002FLongLeo287\u002FOmniClaw\u002Fblob\u002Fmain\u002Fecosystem\u002Fskills\u002Fsglang_diffusion_ako4all_kernel\u002Fschema.json",{"path":88,"resourceType":89,"resourceTypeDesc":90,"sizeBytes":91,"sourceHtmlUrl":92},"scripts\u002Fensure_ako4all_clean.sh","SCRIPT","脚本",2456,"https:\u002F\u002Fgithub.com\u002FLongLeo287\u002FOmniClaw\u002Fblob\u002Fmain\u002Fecosystem\u002Fskills\u002Fsglang_diffusion_ako4all_kernel\u002Fscripts\u002Fensure_ako4all_clean.sh",[94,133,154,181,197,224],{"skillNo":95,"name":96,"nameZh":6,"description":97,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":98,"categories":99,"keywords":112,"publishedAt":41},"01M6JR0KQP4MHT32TN40D9CBFX","UI-UX Pro Max","Advanced UI\u002FUX analysis and generation skills with data-driven patterns. Python scripts (core.py, search.py) for systematic UI pattern extraction, design system generation, and component search from curated data sources.","https:\u002F\u002Fgithub.com\u002FLongLeo287\u002FOmniClaw\u002Fblob\u002Fmain\u002Fecosystem\u002Fui_components\u002Fui_ux_pro_max\u002FSKILL.md",[100,103,104,109],{"categoryNo":101,"categoryName":102,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000015","视觉设计",{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},{"categoryNo":105,"categoryName":106,"taxonomyType":107,"taxonomyTypeDesc":108},"01JBSEEDCAT000000000000032","设计师","PROFESSION","适用职业",{"categoryNo":110,"categoryName":111,"taxonomyType":107,"taxonomyTypeDesc":108},"01JBSEEDCAT000000000000033","数据分析师",[113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132],"ux","ui","ui ux pro max","max","pro","ui ux","design","component","search","analysis","team","rd","ui-ux","active","systematic","scripts","pattern extraction","sources","advanced","data-driven",{"skillNo":134,"name":135,"nameZh":6,"description":136,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":137,"categories":138,"keywords":140,"publishedAt":41},"01M6JR0EF5AEQHP2FSTB57H05P","shadcn-ui","Expert guidance for integrating and building applications with shadcn\u002Fui components, including component discovery, installation, customization, and best practices.","https:\u002F\u002Fgithub.com\u002FLongLeo287\u002FOmniClaw\u002Fblob\u002Fmain\u002Fecosystem\u002Fui_components\u002Fshadcn_ui\u002FSKILL.md",[139],{"categoryNo":105,"categoryName":106,"taxonomyType":107,"taxonomyTypeDesc":108},[141,114,142,143,120,144,145,146,147,123,148,149,150,126,151,152,31,153],"shadcn","shadcn ui","components","core","best practices","applications","discovery","customization","guidance","expert","installation","ui components","for",{"skillNo":155,"name":156,"nameZh":6,"description":157,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":158,"categories":159,"keywords":163,"publishedAt":41},"01M6JR08F51BEN17RHFYX65WVE","zustand_store_ts","Create Zustand stores with TypeScript, subscribeWithSelector middleware, and proper state\u002Faction separation. Use when building React state management, creating global stores, or implementing reactive state patterns with Zustand.","https:\u002F\u002Fgithub.com\u002FLongLeo287\u002FOmniClaw\u002Fblob\u002Fmain\u002Fecosystem\u002Fskills\u002Fzustand_store_ts\u002FSKILL.md",[160],{"categoryNo":161,"categoryName":162,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000009","前端开发",[164,165,166,167,168,169,170,171,172,173,174,175,35,176,177,178,31,179,180],"zustand store ts","zustand","store","typescript","microsoft","react","create","state management","state","react state management","management","global","patterns","separation","middleware","action","implementing",{"skillNo":182,"name":165,"nameZh":6,"description":183,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":184,"categories":185,"keywords":186,"publishedAt":41},"01M6JR00TDMVAFFCZXZ04AN7Q7","Zustand state management guide. Use when working with store code (src\u002Fstore\u002F**), implementing actions, managing state, or creating slices. Triggers on Zustand store development, state management questions, or action implementation.","https:\u002F\u002Fgithub.com\u002FLongLeo287\u002FOmniClaw\u002Fblob\u002Fmain\u002Fecosystem\u002Fskills\u002Fzustand\u002FSKILL.md",[],[165,187,171,188,172,189,190,174,191,35,192,193,194,166,31,195,196,179,180],"development","code","actions","src","guide","implementation","zustand state","triggers","questions","zustand state management",{"skillNo":198,"name":199,"nameZh":6,"description":200,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":201,"categories":202,"keywords":207,"publishedAt":41},"01M6JQZSWV4KWJP112G6XHEP44","youtube_summarizer","Automatically fetch YouTube video transcripts, generate structured summaries, and send full transcripts to messaging platforms. Detects YouTube URLs and provides metadata, key insights, and downloadable transcripts.","https:\u002F\u002Fgithub.com\u002FLongLeo287\u002FOmniClaw\u002Fblob\u002Fmain\u002Fecosystem\u002Fskills\u002Fyoutube_summarizer\u002FSKILL.md",[203,206],{"categoryNo":204,"categoryName":205,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000016","图像与多媒体",{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},[208,209,210,211,212,213,214,215,216,217,218,219,220,221,222,223],"youtube","youtube summarizer","summarizer","generate","metadata","messaging","Video","fetch","telegram","transcription","summarization","structured","full","insights","transcripts","summaries",{"skillNo":225,"name":226,"nameZh":6,"description":227,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":228,"categories":229,"keywords":230,"publishedAt":41},"01M6JQZN3QJB6B77JGKJD62DQT","youtube_clipper","YouTube 视频智能剪辑工具。下载视频和字幕，AI 分析生成精细章节（几分钟级别）， 用户选择片段后自动剪辑、翻译字幕为中英双语、烧录字幕到视频，并生成总结文案。 使用场景：当用户需要剪辑 YouTube 视频、生成短视频片段、制作双语字幕版本时。 关键词：视频剪辑、YouTube、字幕翻译、双语字幕、视频下载、clip video","https:\u002F\u002Fgithub.com\u002FLongLeo287\u002FOmniClaw\u002Fblob\u002Fmain\u002Fecosystem\u002Fskills\u002Fyoutube_clipper\u002FSKILL.md",[],[208,231,232,214,233,234,235,236,237,238,239,240,241,242,243,244,245],"clipper","youtube clipper","Bash","下载","视频","生成","关键词","翻译","read","write","clip","总结","工具","文案","sonnet"]