[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6XCT12A9K71NQ7SQMBPB018":3,"skill-same:01M6XCT12A9K71NQ7SQMBPB018":68},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":36,"linkOnly":37,"markdownContent":38,"markdownBody":39,"frontMatter":40,"refName":41,"commitSha":42,"skillDirectory":43,"forksCount":13,"repositoryHtmlUrl":44,"repositoryHomepageUrl":6,"ownerHtmlUrl":45,"ownerBlogUrl":6,"ownerDescription":46,"updatedAt":47,"install":48,"resources":61},"01M6XCT12A9K71NQ7SQMBPB018","upload-parity-experiments",null,"Create or reuse Hugging Face dataset PRs for `harborframework\u002Fparity-experiments` and upload Harbor parity\u002Foracle result folders efficiently with sparse checkout, raw git pushes, and Git LFS.","Waaaalnut\u002Fskillsbench","Waaaalnut","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F103995656?v=4","01M6XCS9K1V51EY3JMB7DGMX24","01M6XCS9Q4WPW88W62C55G1H44",0,"Apache-2.0","https:\u002F\u002Fgithub.com\u002FWaaaalnut\u002Fskillsbench\u002Fblob\u002Fmain\u002Fharbor\u002Fskills\u002Fupload-parity-experiments\u002FSKILL.md",[],[18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35],"experiments","parity","upload","upload parity experiments","face","oracle","result","prs","for","lfs","with","reuse","checkout","dataset","hugging face","harbor","git","create","2026-11-10 10:12:30",false,"---\nname: upload-parity-experiments\ndescription: Create or reuse Hugging Face dataset PRs for `harborframework\u002Fparity-experiments` and upload Harbor parity\u002Foracle result folders efficiently with sparse checkout, raw git pushes, and Git LFS.\n---\n\n# Upload Parity Experiments\n\nUse this skill to publish Harbor parity experiment outputs to the shared Hugging Face dataset and capture the resulting discussion URL for the adapter's `parity_pr` field.\n\n## Why This Skill Exists\n\n- `hf upload-large-folder` can be slow or unreliable for large parity bundles because it pushes through the Hub API commit loop.\n- A normal git clone of `harborframework\u002Fparity-experiments` is too expensive because the dataset is very large.\n- Hugging Face dataset PR refs are different from GitHub PR refs and are easy to misuse.\n- Files larger than 10 MiB must be Git LFS-tracked before pushing.\n\nThis skill avoids the full clone by fetching only the target PR ref with `--depth 1 --filter=blob:none` and checking out only the paths needed for the current adapter.\n\n## Prereqs\n\n- Ensure Hugging Face authentication is already available. Prefer `hf auth login`.\n- Keep the target dataset fixed to `harborframework\u002Fparity-experiments` unless the user explicitly asks for another repo.\n- Accept any local upload source that already contains the final files the user wants to publish.\n\n## Preferred Workflow\n\n1. Create or reuse a dataset PR.\n2. Prepare a sparse local worktree for that PR ref.\n3. Copy the local parity results into the sparse checkout.\n4. Ensure every file larger than 10 MiB is Git LFS-tracked before committing.\n5. Push directly to the PR ref with raw `git push`.\n6. Share the discussion URL and record it as the adapter's `parity_pr`.\n\nFor large parity bundles, prefer raw git over `hf upload-large-folder`. The raw git path is materially faster and more reliable because it avoids the API-side commit loop and does not require cloning the entire parity dataset.\n\n## 1. Create Or Reuse A Dataset PR\n\nIf the user already has a parity PR number, reuse it.\n\nOtherwise, create one with the bundled helper:\n\n```bash\nuv run python scripts\u002Fcreate_pr.py create-pr \\\n  --title \"Add parity experiments for \u003Cadapter_name>\" \\\n  --description-file \u002Fpath\u002Fto\u002Fpr-description.md\n```\n\nThe script prints JSON including:\n\n- `pr_number`\n- `discussion_url`\n- `repo_id`\n\n## 2. Prepare A Sparse PR Checkout\n\n```bash\nmkdir -p \u002Ftmp\u002Fparity-experiments-pr\u003Cnumber>\ncd \u002Ftmp\u002Fparity-experiments-pr\u003Cnumber>\n\ngit init\ngit remote add origin git@hf.co:datasets\u002Fharborframework\u002Fparity-experiments\ngit config core.sparseCheckout true\ngit sparse-checkout init --cone\ngit sparse-checkout set .gitattributes adapters\u002F\u003Cadapter_name>\n\ngit fetch --depth 1 --filter=blob:none origin refs\u002Fpr\u002F\u003Cnumber>:pr\u002F\u003Cnumber>\ngit checkout pr\u002F\u003Cnumber>\n```\n\nThis fetches only the PR ref and the requested paths instead of cloning the full dataset repo.\n\n## 3. Copy The Local Results\n\nIf the local folder already contains the final repo-root layout, copy it as-is.\n\nIf the local folder only contains the adapter subtree, copy it into `adapters\u002F\u003Cadapter_name>\u002F`:\n\n```bash\nrsync -a --delete \\\n  --exclude '.git' \\\n  --exclude '.cache' \\\n  --exclude '.DS_Store' \\\n  \u002Fpath\u002Fto\u002Flocal-folder\u002F \\\n  adapters\u002F\u003Cadapter_name>\u002F\n```\n\n## 4. Ensure Large Files Use Git LFS\n\nScan the sparse checkout for files larger than 10 MiB:\n\n```bash\npython - \u003C\u003C'PY'\nfrom pathlib import Path\n\nfor path in sorted(Path(\".\").rglob(\"*\")):\n    if path.is_file() and \".git\" not in path.parts and path.stat().st_size > 10 * 1024 * 1024:\n        print(path)\nPY\n```\n\nIf any file is larger than 10 MiB, make sure it is LFS-tracked before committing:\n\n```bash\ngit lfs track \"adapters\u002F\u003Cadapter_name>\u002F\u003Cpattern>\"\ngit add .gitattributes\n```\n\n## 5. Commit And Push\n\n```bash\nfind . -name .DS_Store -delete\ngit add .gitattributes adapters\u002F\u003Cadapter_name>\ngit commit -m \"Add parity experiment artifacts for \u003Cadapter_name>\"\nGIT_SSH_COMMAND='ssh -o ServerAliveInterval=30 -o ServerAliveCountMax=10' \\\n  git push origin pr\u002F\u003Cnumber>:refs\u002Fpr\u002F\u003Cnumber>\n```\n\n## 6. Record The Discussion URL\n\nUse this discussion URL in `parity_experiment.json`:\n\n```text\nhttps:\u002F\u002Fhuggingface.co\u002Fdatasets\u002Fharborframework\u002Fparity-experiments\u002Fdiscussions\u002F\u003Cnumber>\n```\n\n## Monitoring And Retry\n\nWhen raw `git push` is running, check the live output first. Large pushes often spend a long time in:\n\n- `git-lfs pre-push`\n- single-object LFS uploads\n- packfile compression\n\nUseful checks:\n\n```bash\ngit ls-remote git@hf.co:datasets\u002Fharborframework\u002Fparity-experiments refs\u002Fpr\u002F\u003Cnumber>\n```\n\n```bash\nps -p \u003Cpid> -o pid=,etime=,command=\n```\n\nIf the push exits early:\n\n- rerun the exact same `git push` command first\n- if Hugging Face rejects a file larger than 10 MiB, add the needed `git lfs track` rules, re-add the file, and retry\n- if the connection broke mid-push, rerun the same push command before changing anything else\n\n## Fallback\n\nUse `hf upload-large-folder` only as a fallback when raw git is unavailable or the user explicitly asks for it.\n\nYou can still create the PR with the bundled helper and upload to the PR revision:\n\n```bash\nhf upload-large-folder harborframework\u002Fparity-experiments \\\n  \u003Clocal-folder> \\\n  --repo-type dataset \\\n  --revision refs\u002Fpr\u002F\u003Cpr-number> \\\n  --exclude \".DS_Store\" \\\n  --exclude \"**\u002F.DS_Store\" \\\n  --num-workers 8\n```\n\n## Guardrails\n\n- Do not upload the Harbor repo itself by accident. Upload only the intended local results folder.\n- Do not fall back to a full clone of `harborframework\u002Fparity-experiments`.\n- Always remove or exclude `.DS_Store`.\n- Before pushing, make sure every file larger than 10 MiB is LFS-tracked.\n- If the user already has a parity PR number, reuse it instead of creating another one.\n","# Upload Parity Experiments\n\nUse this skill to publish Harbor parity experiment outputs to the shared Hugging Face dataset and capture the resulting discussion URL for the adapter's `parity_pr` field.\n\n## Why This Skill Exists\n\n- `hf upload-large-folder` can be slow or unreliable for large parity bundles because it pushes through the Hub API commit loop.\n- A normal git clone of `harborframework\u002Fparity-experiments` is too expensive because the dataset is very large.\n- Hugging Face dataset PR refs are different from GitHub PR refs and are easy to misuse.\n- Files larger than 10 MiB must be Git LFS-tracked before pushing.\n\nThis skill avoids the full clone by fetching only the target PR ref with `--depth 1 --filter=blob:none` and checking out only the paths needed for the current adapter.\n\n## Prereqs\n\n- Ensure Hugging Face authentication is already available. Prefer `hf auth login`.\n- Keep the target dataset fixed to `harborframework\u002Fparity-experiments` unless the user explicitly asks for another repo.\n- Accept any local upload source that already contains the final files the user wants to publish.\n\n## Preferred Workflow\n\n1. Create or reuse a dataset PR.\n2. Prepare a sparse local worktree for that PR ref.\n3. Copy the local parity results into the sparse checkout.\n4. Ensure every file larger than 10 MiB is Git LFS-tracked before committing.\n5. Push directly to the PR ref with raw `git push`.\n6. Share the discussion URL and record it as the adapter's `parity_pr`.\n\nFor large parity bundles, prefer raw git over `hf upload-large-folder`. The raw git path is materially faster and more reliable because it avoids the API-side commit loop and does not require cloning the entire parity dataset.\n\n## 1. Create Or Reuse A Dataset PR\n\nIf the user already has a parity PR number, reuse it.\n\nOtherwise, create one with the bundled helper:\n\n```bash\nuv run python scripts\u002Fcreate_pr.py create-pr \\\n  --title \"Add parity experiments for \u003Cadapter_name>\" \\\n  --description-file \u002Fpath\u002Fto\u002Fpr-description.md\n```\n\nThe script prints JSON including:\n\n- `pr_number`\n- `discussion_url`\n- `repo_id`\n\n## 2. Prepare A Sparse PR Checkout\n\n```bash\nmkdir -p \u002Ftmp\u002Fparity-experiments-pr\u003Cnumber>\ncd \u002Ftmp\u002Fparity-experiments-pr\u003Cnumber>\n\ngit init\ngit remote add origin git@hf.co:datasets\u002Fharborframework\u002Fparity-experiments\ngit config core.sparseCheckout true\ngit sparse-checkout init --cone\ngit sparse-checkout set .gitattributes adapters\u002F\u003Cadapter_name>\n\ngit fetch --depth 1 --filter=blob:none origin refs\u002Fpr\u002F\u003Cnumber>:pr\u002F\u003Cnumber>\ngit checkout pr\u002F\u003Cnumber>\n```\n\nThis fetches only the PR ref and the requested paths instead of cloning the full dataset repo.\n\n## 3. Copy The Local Results\n\nIf the local folder already contains the final repo-root layout, copy it as-is.\n\nIf the local folder only contains the adapter subtree, copy it into `adapters\u002F\u003Cadapter_name>\u002F`:\n\n```bash\nrsync -a --delete \\\n  --exclude '.git' \\\n  --exclude '.cache' \\\n  --exclude '.DS_Store' \\\n  \u002Fpath\u002Fto\u002Flocal-folder\u002F \\\n  adapters\u002F\u003Cadapter_name>\u002F\n```\n\n## 4. Ensure Large Files Use Git LFS\n\nScan the sparse checkout for files larger than 10 MiB:\n\n```bash\npython - \u003C\u003C'PY'\nfrom pathlib import Path\n\nfor path in sorted(Path(\".\").rglob(\"*\")):\n    if path.is_file() and \".git\" not in path.parts and path.stat().st_size > 10 * 1024 * 1024:\n        print(path)\nPY\n```\n\nIf any file is larger than 10 MiB, make sure it is LFS-tracked before committing:\n\n```bash\ngit lfs track \"adapters\u002F\u003Cadapter_name>\u002F\u003Cpattern>\"\ngit add .gitattributes\n```\n\n## 5. Commit And Push\n\n```bash\nfind . -name .DS_Store -delete\ngit add .gitattributes adapters\u002F\u003Cadapter_name>\ngit commit -m \"Add parity experiment artifacts for \u003Cadapter_name>\"\nGIT_SSH_COMMAND='ssh -o ServerAliveInterval=30 -o ServerAliveCountMax=10' \\\n  git push origin pr\u002F\u003Cnumber>:refs\u002Fpr\u002F\u003Cnumber>\n```\n\n## 6. Record The Discussion URL\n\nUse this discussion URL in `parity_experiment.json`:\n\n```text\nhttps:\u002F\u002Fhuggingface.co\u002Fdatasets\u002Fharborframework\u002Fparity-experiments\u002Fdiscussions\u002F\u003Cnumber>\n```\n\n## Monitoring And Retry\n\nWhen raw `git push` is running, check the live output first. Large pushes often spend a long time in:\n\n- `git-lfs pre-push`\n- single-object LFS uploads\n- packfile compression\n\nUseful checks:\n\n```bash\ngit ls-remote git@hf.co:datasets\u002Fharborframework\u002Fparity-experiments refs\u002Fpr\u002F\u003Cnumber>\n```\n\n```bash\nps -p \u003Cpid> -o pid=,etime=,command=\n```\n\nIf the push exits early:\n\n- rerun the exact same `git push` command first\n- if Hugging Face rejects a file larger than 10 MiB, add the needed `git lfs track` rules, re-add the file, and retry\n- if the connection broke mid-push, rerun the same push command before changing anything else\n\n## Fallback\n\nUse `hf upload-large-folder` only as a fallback when raw git is unavailable or the user explicitly asks for it.\n\nYou can still create the PR with the bundled helper and upload to the PR revision:\n\n```bash\nhf upload-large-folder harborframework\u002Fparity-experiments \\\n  \u003Clocal-folder> \\\n  --repo-type dataset \\\n  --revision refs\u002Fpr\u002F\u003Cpr-number> \\\n  --exclude \".DS_Store\" \\\n  --exclude \"**\u002F.DS_Store\" \\\n  --num-workers 8\n```\n\n## Guardrails\n\n- Do not upload the Harbor repo itself by accident. Upload only the intended local results folder.\n- Do not fall back to a full clone of `harborframework\u002Fparity-experiments`.\n- Always remove or exclude `.DS_Store`.\n- Before pushing, make sure every file larger than 10 MiB is LFS-tracked.\n- If the user already has a parity PR number, reuse it instead of creating another one.\n",{"name":5,"description":7},"main","7eeb258ada72b7f58119b0e7a2c00ea89d9beedd","harbor\u002Fskills\u002Fupload-parity-experiments","https:\u002F\u002Fgithub.com\u002FWaaaalnut\u002Fskillsbench","https:\u002F\u002Fgithub.com\u002FWaaaalnut","student of Zhejiang University","2026-04-21 23:16:23",{"source":8,"skillName":5,"commands":49,"prompt":59,"zipAvailable":60},[50,53,56],{"tool":51,"command":52},"npx","npx skills add Waaaalnut\u002Fskillsbench --skill upload-parity-experiments",{"tool":54,"command":55},"bunx","bunx skills add Waaaalnut\u002Fskillsbench --skill upload-parity-experiments",{"tool":57,"command":58},"pnpm","pnpm dlx skills add Waaaalnut\u002Fskillsbench --skill upload-parity-experiments","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FWaaaalnut\u002Fskillsbench\u002Fblob\u002Fmain\u002Fharbor\u002Fskills\u002Fupload-parity-experiments\u002FSKILL.md",true,[62],{"path":63,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":66,"sourceHtmlUrl":67},"scripts\u002Fcreate_pr.py","SCRIPT","脚本",2171,"https:\u002F\u002Fgithub.com\u002FWaaaalnut\u002Fskillsbench\u002Fblob\u002Fmain\u002Fharbor\u002Fskills\u002Fupload-parity-experiments\u002Fscripts\u002Fcreate_pr.py",[69,83,98,125,147,165],{"skillNo":70,"name":71,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":72,"categories":73,"keywords":79,"publishedAt":36},"01M6XEWMJ6VY3ZC53MPDCPG3GH","data-reconciliation","https:\u002F\u002Fgithub.com\u002FWaaaalnut\u002Fskillsbench\u002Fblob\u002Fmain\u002Ftasks\u002Fxlsx-recover-data\u002Fenvironment\u002Fskills\u002Fdata-reconciliation\u002FSKILL.md",[74],{"categoryNo":75,"categoryName":76,"taxonomyType":77,"taxonomyTypeDesc":78},"01JBSEEDCAT000000000000033","数据分析师","PROFESSION","适用职业",[80,81,82],"data reconciliation","reconciliation","data",{"skillNo":84,"name":85,"nameZh":6,"description":86,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":87,"categories":88,"keywords":89,"publishedAt":36},"01M6XEW8MNPMF1KRXV6ATGFRQF","speech-to-text","Transcribe video to timestamped text using Whisper tiny model (pre-installed).","https:\u002F\u002Fgithub.com\u002FWaaaalnut\u002Fskillsbench\u002Fblob\u002Fmain\u002Ftasks\u002Fvideo-tutorial-indexer\u002Fenvironment\u002Fskills\u002Fspeech-to-text\u002FSKILL.md",[],[90,91,92,93,94,95,96,97],"speech","text","speech to text","transcribe video","whisper","model","Video","transcribe",{"skillNo":99,"name":100,"nameZh":6,"description":101,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":102,"categories":103,"keywords":109,"publishedAt":36},"01M6XEW3FQZP0T1RKQQJFXVY9T","video-processor","Process videos by removing segments and concatenating remaining parts. Use when you need to remove detected pauses\u002Fopenings from videos, create highlight reels, or batch process segment removals using ffmpeg filter_complex.","https:\u002F\u002Fgithub.com\u002FWaaaalnut\u002Fskillsbench\u002Fblob\u002Fmain\u002Ftasks\u002Fvideo-silence-remover\u002Fenvironment\u002Fskills\u002Fvideo-processor\u002FSKILL.md",[104],{"categoryNo":105,"categoryName":106,"taxonomyType":107,"taxonomyTypeDesc":108},"01JBSEEDCAT000000000000016","图像与多媒体","SCENARIO","使用场景",[96,110,111,112,35,113,114,115,116,117,118,119,120,121,122,123,124],"processor","video processor","batch","reels","when","parts","segments","process","videos","you","segment","filter","ffmpeg","use","remove",{"skillNo":126,"name":127,"nameZh":6,"description":128,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":129,"categories":130,"keywords":132,"publishedAt":36},"01M6XEVXJ5KZ0NTY3XZKRY4AS1","silence-detector","Detect initial silence segments in audio\u002Fvideo using energy-based analysis. Use when you need to find low-energy periods at the start of recordings (title slides, setup time, pre-roll silence).","https:\u002F\u002Fgithub.com\u002FWaaaalnut\u002Fskillsbench\u002Fblob\u002Fmain\u002Ftasks\u002Fvideo-silence-remover\u002Fenvironment\u002Fskills\u002Fsilence-detector\u002FSKILL.md",[131],{"categoryNo":105,"categoryName":106,"taxonomyType":107,"taxonomyTypeDesc":108},[133,134,135,136,114,119,137,138,139,140,141,116,142,143,96,144,145,146,123],"detector","silence detector","title","slides","time","energy","at","recordings","detect","find","Audio","analysis","start","setup",{"skillNo":148,"name":149,"nameZh":6,"description":150,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":151,"categories":152,"keywords":153,"publishedAt":36},"01M6XEVT3NGRN1FT7JPMAPXS6M","segment-combiner","Combine multiple segment detection results into a unified list. Use when you need to merge segments from different detectors, prepare removal lists for video processing, or consolidate detection outputs.","https:\u002F\u002Fgithub.com\u002FWaaaalnut\u002Fskillsbench\u002Fblob\u002Fmain\u002Ftasks\u002Fvideo-silence-remover\u002Fenvironment\u002Fskills\u002Fsegment-combiner\u002FSKILL.md",[],[120,154,96,155,156,157,158,26,114,159,160,116,119,161,162,123,163,164],"segment combiner","detection","prepare","merge","video processing","outputs","combine","processing","consolidate","results","list",{"skillNo":166,"name":167,"nameZh":6,"description":168,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":169,"categories":170,"keywords":174,"publishedAt":36},"01M6XEVMPVASGS2KX86QK4FMT7","report-generator","Generate compression reports for video processing. Use when you need to create structured JSON reports with duration statistics, compression ratios, and segment details after video processing.","https:\u002F\u002Fgithub.com\u002FWaaaalnut\u002Fskillsbench\u002Fblob\u002Fmain\u002Ftasks\u002Fvideo-silence-remover\u002Fenvironment\u002Fskills\u002Freport-generator\u002FSKILL.md",[171],{"categoryNo":172,"categoryName":173,"taxonomyType":77,"taxonomyTypeDesc":78},"01JBSEEDCAT000000000000034","行政与办公人员",[175,176,177,120,114,26,119,28,161,178,179,180,123,158,181,96,182,183,35],"generator","report generator","report","statistics","structured","reports","json","compression","generate"]