[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill-nvidia-nvidia-skill-finder":3,"mdc--3cobg3-key":28,"related-org-nvidia-nvidia-skill-finder":715,"related-repo-nvidia-nvidia-skill-finder":876},{"slug":4,"name":4,"fn":5,"description":6,"org":7,"tags":11,"stars":17,"repoUrl":18,"updatedAt":19,"license":20,"forks":21,"topics":22,"repo":23,"sourceUrl":26,"mdContent":27},"nvidia-skill-finder","find relevant NVIDIA skills","Use for NVIDIA-related requests where an NVIDIA skill might help, even if the user did not ask for a skill. Trigger on NVIDIA products, hardware, software, SDKs, GPUs, Jetson\u002FJetPack\u002FL4T\u002FBSP\u002FSDK Manager\u002Fdriver\u002Fflashing\u002Fsetup, CUDA, NIM, NeMo, Omniverse\u002FOpenUSD\u002FSimReady, RAPIDS\u002FcuDF, cuPyNumeric, cuOpt, Dynamo, Holoscan, TensorRT, DeepStream, VSS, TAO, NGC\u002FNVCF. Do not use for generic non-NVIDIA route, optimize, deploy, AI, video, data, or infrastructure tasks.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},"nvidia","NVIDIA","https:\u002F\u002Fpexgzepcugksgbtrxkhf.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Forg-logos\u002Fnvidia.png",[12,16],{"name":13,"slug":14,"type":15},"Reference","reference","tag",{"name":9,"slug":8,"type":15},2473,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Fskills","2026-07-17T06:07:12.467436","CC-BY-4.0 AND Apache-2.0",281,[],{"repoUrl":18,"stars":17,"forks":21,"topics":24,"description":25},[],"AI agent skills published by NVIDIA","https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Fskills\u002Ftree\u002FHEAD\u002Fskills\u002Fnvidia-skill-finder","---\nname: nvidia-skill-finder\ndescription: >-\n  Use for NVIDIA-related requests where an NVIDIA skill might help, even if the user did not ask for a skill. Trigger on NVIDIA products, hardware, software, SDKs, GPUs, Jetson\u002FJetPack\u002FL4T\u002FBSP\u002FSDK Manager\u002Fdriver\u002Fflashing\u002Fsetup, CUDA, NIM, NeMo, Omniverse\u002FOpenUSD\u002FSimReady, RAPIDS\u002FcuDF, cuPyNumeric, cuOpt, Dynamo, Holoscan, TensorRT, DeepStream, VSS, TAO, NGC\u002FNVCF. Do not use for generic non-NVIDIA route, optimize, deploy, AI, video, data, or infrastructure tasks.\nlicense: CC-BY-4.0 AND Apache-2.0\nmetadata:\n  author: NVIDIA\n  tags:\n    - nvidia\n    - skills\n    - discovery\n    - catalog\n    - router\n  domain: agent-skills\n---\n\n# NVIDIA Skill Finder\n\n## Purpose\n\nHelp users discover, install, and start using NVIDIA skills that may not be\ninstalled yet. Treat this skill as a stable NVIDIA capability detector and\ncatalog router, not as a mirror of every external skill's trigger text.\n\nUse the live catalog as the source of truth for specific skill names,\ndescriptions, and availability. Keep only stable taxonomy guidance here.\n\n## When to Use this Skill \nUse this skill to find the best NVIDIA skill for a product, task, or workflow.\nThe user does not need to explicitly ask for a skill. If the request is about\nNVIDIA hardware, software, SDKs, drivers, setup, troubleshooting, or an\nNVIDIA-adjacent workflow, check whether the live catalog has a skill that could\nhelp before proceeding too far with general guidance.\n\nTypical triggers:\n\n- Asks \"how do I do X\" where X might be a common task that an NVIDIA skill can help with.\n- Says \"find a skill for X\" or \"is there a skill for X\".\n- Expresses interest in extending agent capabilities in NVIDIA domains.\n- Mentions they need help with a specific NVIDIA catalog domain covered below.\n- Asks about installing, configuring, troubleshooting, or using an NVIDIA\n  product, device, SDK, or service.\n\nContinue with this skill only when the request is plausibly related to an NVIDIA product area or taxonomy category.\n\nStrong signals:\n\n- The user mentions NVIDIA, CUDA, GPU acceleration, NIM, NeMo, Omniverse, OpenUSD,\n  SimReady, cuOpt, RAPIDS\u002FcuDF, cuPyNumeric, Dynamo, Holoscan, TensorRT, VSS,\n  DeepStream, Jetson, JetPack, L4T, BSP, SDK Manager, TAO, NGC, NVCF, or another\n  NVIDIA product.\n- The task maps strongly to an NVIDIA catalog lane such as Agentic AI, Physical\n  AI, Robotics, Vision AI, Conversational AI, Simulation and Modeling, Data\n  Science, Training AI, Inference AI, Decision Optimization, GPU Development,\n  Quantum Computing, Infrastructure, or Networking.\n- The task uses distinctive phrases such as RAG\u002Fdeep research, vehicle routing,\n  LP\u002FMILP\u002FQP, GPU DataFrames, multi-GPU NumPy\u002FSciPy, KV-aware routing,\n  Jetson driver install, JetPack flashing, BSP download, SDK Manager setup,\n  CAD-to-SimReady, OpenUSD optimization, VSS\u002Fvideo search\u002Fsummarization, DICOM\n  workflows, robotics simulation, Holoscan setup, or synthetic data generation.\n\nRead [references\u002Ftaxonomy-routing.md](references\u002Ftaxonomy-routing.md) only when\nthe request is taxonomy-only, ambiguous, or needs browse\u002Fdomain mapping. For\nobvious product-name matches, go directly to live catalog lookup.\n\n## Implicit Invocation Constraints\n\nImplicit invocation is intentional: this skill acts as a NVIDIA capability\ndetector and catalog router. This scopes the skill to NVIDIA relevance; it does\nnot narrow it. Use the full trigger breadth in \"When to Use this Skill\" —\nincluding the softer \"how do I do X\" triggers and any request plausibly related\nto an NVIDIA product area or catalog taxonomy lane.\n\nThe gate is relevance, not consent. Do not activate for the generic software\ntasks in \"When Not to Use this Skill\" unless they also carry an NVIDIA, GPU,\naccelerated-computing, or distinctive NVIDIA workflow signal.\n\nRecommending a skill is always allowed once the request is relevant. Installing\nor modifying skills is not: never run an install (e.g. `npx skills add`) or\nchange agent capabilities without explicit user approval. A catalog match is\nonly a recommendation until the user confirms.\n\n## When Not to Use this Skill\nStay quiet when the request is generic:\n\n- \"route\" means an HTTP route, Express route, file route, or request routing.\n- \"optimize\" means ordinary web performance, CSS, bundle size, SQL tuning, or\n  generic code cleanup.\n- \"deploy\" means generic Kubernetes, cloud, CI\u002FCD, or web hosting.\n- \"AI\", \"data science\", or \"infrastructure\" appears without NVIDIA, GPU,\n  accelerated-computing, or one of the distinctive intent signals above.\n- \"video\" means ordinary trimming, captions, export, or social-media editing.\n\nIf relevance is uncertain, do not interrupt the user's main task. Mention the\nNVIDIA catalog only as an optional aside after answering, or ask one concise\nclarifying question if the choice materially changes the work.\n\n## Instructions\nHow to Help Users Find Skills - a Discovery Workflow\n\nThis skill's first job is skill discovery. For NVIDIA-related requests, do a\ncatalog check before using general web search, NVIDIA product docs, or general\nproduct knowledge as the main answer. Product documentation can help after the\ncatalog check, but it is not a substitute for checking the NVIDIA skills catalog.\n\nCatalog check means one of:\n\n- `npx skills add nvidia\u002Fskills --list`\n- `https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Fskills\u002Ftree\u002Fmain\u002Fskills`\n- `https:\u002F\u002Fbuild.nvidia.com\u002Fskills`\n- `https:\u002F\u002Fraw.githubusercontent.com\u002FNVIDIA\u002Fskills\u002Fmain\u002Fskills.sh.json`\n\n1. Check whether the relevant NVIDIA skill is already installed or already in context. If it is, hand off to that skill instead of recommending install.\n2. Query the live catalog before naming a specific install target or giving the\n   main product answer. When shell access is available, attempt this command first:\n\n```bash\nnpx skills add nvidia\u002Fskills --list\n```\n\nUse the fallback catalog sources only if the CLI is unavailable, blocked, or fails:\n- https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Fskills\u002Ftree\u002Fmain\u002Fskills\n- https:\u002F\u002Fbuild.nvidia.com\u002Fskills\n- https:\u002F\u002Fraw.githubusercontent.com\u002FNVIDIA\u002Fskills\u002Fmain\u002Fskills.sh.json\n\nDo not count a general web search, developer.nvidia.com product documentation,\nor docs.nvidia.com product documentation as the catalog check.\n\n3. Match the request against current skill names, descriptions, product groups, and skill cards. Prefer NVIDIA-verified catalog entries over memory.\n4. If a current catalog skill strongly matches, recommend it before continuing\n   with general product guidance. Recommend at most three skills, ordered by\n   confidence. For each, include: skill name, why it fits, install command, and\n   first useful prompt.\n5. Ask before installing. Do not run `npx skills add` unless the user approves.\n\n## Recommendation Format\n\nFor a strong match that is not installed:\n\n```text\nThe NVIDIA \u003Cproduct or skill family> skill could help with this. Would you like me to install \u003Cskill-name>?\n```\n\nUse the active agent target when known:\n\n```bash\nnpx skills add nvidia\u002Fskills --skill \u003Cskill-name> --agent codex --global --yes\nnpx skills add nvidia\u002Fskills --skill \u003Cskill-name> --agent claude-code --global --yes\n```\n\nIf the agent target is unknown, omit `--agent` and let the CLI prompt:\n\n```bash\nnpx skills add nvidia\u002Fskills --skill \u003Cskill-name> --global --yes\n```\n\nAfter install, tell the user to restart or reload the agent if their client\ndoes not pick up newly installed skills immediately.\n\n## Confidence Rules\n\n- High confidence: product name or distinctive intent maps cleanly to one\n  current catalog skill or skill family.\n- Medium confidence: taxonomy lane matches but several skills may apply. Offer\n  a short shortlist and ask which direction matches the user's task.\n- Low confidence: only generic wording matches. Do the user's task without\n  recommending an NVIDIA skill.\n\nDo not fabricate catalog entries or guess at skill names. If catalog lookup\nfails, say the catalog could not be checked and do not name a concrete slug or\nemit an `npx skills add ... --skill \u003Cname>` install command. Offer to share the\ncatalog URL, retry lookup, or continue with general help.\n\n## When No Skills are Found\n\nSay that no strong NVIDIA catalog match was found, then either:\n1. search the broader ecosystem with npx skills find \u003Cquery>, or\n2. offer to help directly without a skill, or\n3. suggest creating a new skill if the workflow is recurring.\n",{"data":29,"body":37},{"name":4,"description":6,"license":20,"metadata":30},{"author":9,"tags":31,"domain":36},[8,32,33,34,35],"skills","discovery","catalog","router","agent-skills",{"type":38,"children":39},"root",[40,48,55,61,66,72,77,82,112,117,122,140,153,159,164,169,183,189,194,222,227,233,238,243,248,287,301,345,350,378,383,409,415,420,430,435,565,578,632,637,643,661,674,680,685,709],{"type":41,"tag":42,"props":43,"children":44},"element","h1",{"id":4},[45],{"type":46,"value":47},"text","NVIDIA Skill Finder",{"type":41,"tag":49,"props":50,"children":52},"h2",{"id":51},"purpose",[53],{"type":46,"value":54},"Purpose",{"type":41,"tag":56,"props":57,"children":58},"p",{},[59],{"type":46,"value":60},"Help users discover, install, and start using NVIDIA skills that may not be\ninstalled yet. Treat this skill as a stable NVIDIA capability detector and\ncatalog router, not as a mirror of every external skill's trigger text.",{"type":41,"tag":56,"props":62,"children":63},{},[64],{"type":46,"value":65},"Use the live catalog as the source of truth for specific skill names,\ndescriptions, and availability. Keep only stable taxonomy guidance here.",{"type":41,"tag":49,"props":67,"children":69},{"id":68},"when-to-use-this-skill",[70],{"type":46,"value":71},"When to Use this Skill",{"type":41,"tag":56,"props":73,"children":74},{},[75],{"type":46,"value":76},"Use this skill to find the best NVIDIA skill for a product, task, or workflow.\nThe user does not need to explicitly ask for a skill. If the request is about\nNVIDIA hardware, software, SDKs, drivers, setup, troubleshooting, or an\nNVIDIA-adjacent workflow, check whether the live catalog has a skill that could\nhelp before proceeding too far with general guidance.",{"type":41,"tag":56,"props":78,"children":79},{},[80],{"type":46,"value":81},"Typical triggers:",{"type":41,"tag":83,"props":84,"children":85},"ul",{},[86,92,97,102,107],{"type":41,"tag":87,"props":88,"children":89},"li",{},[90],{"type":46,"value":91},"Asks \"how do I do X\" where X might be a common task that an NVIDIA skill can help with.",{"type":41,"tag":87,"props":93,"children":94},{},[95],{"type":46,"value":96},"Says \"find a skill for X\" or \"is there a skill for X\".",{"type":41,"tag":87,"props":98,"children":99},{},[100],{"type":46,"value":101},"Expresses interest in extending agent capabilities in NVIDIA domains.",{"type":41,"tag":87,"props":103,"children":104},{},[105],{"type":46,"value":106},"Mentions they need help with a specific NVIDIA catalog domain covered below.",{"type":41,"tag":87,"props":108,"children":109},{},[110],{"type":46,"value":111},"Asks about installing, configuring, troubleshooting, or using an NVIDIA\nproduct, device, SDK, or service.",{"type":41,"tag":56,"props":113,"children":114},{},[115],{"type":46,"value":116},"Continue with this skill only when the request is plausibly related to an NVIDIA product area or taxonomy category.",{"type":41,"tag":56,"props":118,"children":119},{},[120],{"type":46,"value":121},"Strong signals:",{"type":41,"tag":83,"props":123,"children":124},{},[125,130,135],{"type":41,"tag":87,"props":126,"children":127},{},[128],{"type":46,"value":129},"The user mentions NVIDIA, CUDA, GPU acceleration, NIM, NeMo, Omniverse, OpenUSD,\nSimReady, cuOpt, RAPIDS\u002FcuDF, cuPyNumeric, Dynamo, Holoscan, TensorRT, VSS,\nDeepStream, Jetson, JetPack, L4T, BSP, SDK Manager, TAO, NGC, NVCF, or another\nNVIDIA product.",{"type":41,"tag":87,"props":131,"children":132},{},[133],{"type":46,"value":134},"The task maps strongly to an NVIDIA catalog lane such as Agentic AI, Physical\nAI, Robotics, Vision AI, Conversational AI, Simulation and Modeling, Data\nScience, Training AI, Inference AI, Decision Optimization, GPU Development,\nQuantum Computing, Infrastructure, or Networking.",{"type":41,"tag":87,"props":136,"children":137},{},[138],{"type":46,"value":139},"The task uses distinctive phrases such as RAG\u002Fdeep research, vehicle routing,\nLP\u002FMILP\u002FQP, GPU DataFrames, multi-GPU NumPy\u002FSciPy, KV-aware routing,\nJetson driver install, JetPack flashing, BSP download, SDK Manager setup,\nCAD-to-SimReady, OpenUSD optimization, VSS\u002Fvideo search\u002Fsummarization, DICOM\nworkflows, robotics simulation, Holoscan setup, or synthetic data generation.",{"type":41,"tag":56,"props":141,"children":142},{},[143,145,151],{"type":46,"value":144},"Read ",{"type":41,"tag":146,"props":147,"children":149},"a",{"href":148},"references\u002Ftaxonomy-routing.md",[150],{"type":46,"value":148},{"type":46,"value":152}," only when\nthe request is taxonomy-only, ambiguous, or needs browse\u002Fdomain mapping. For\nobvious product-name matches, go directly to live catalog lookup.",{"type":41,"tag":49,"props":154,"children":156},{"id":155},"implicit-invocation-constraints",[157],{"type":46,"value":158},"Implicit Invocation Constraints",{"type":41,"tag":56,"props":160,"children":161},{},[162],{"type":46,"value":163},"Implicit invocation is intentional: this skill acts as a NVIDIA capability\ndetector and catalog router. This scopes the skill to NVIDIA relevance; it does\nnot narrow it. Use the full trigger breadth in \"When to Use this Skill\" —\nincluding the softer \"how do I do X\" triggers and any request plausibly related\nto an NVIDIA product area or catalog taxonomy lane.",{"type":41,"tag":56,"props":165,"children":166},{},[167],{"type":46,"value":168},"The gate is relevance, not consent. Do not activate for the generic software\ntasks in \"When Not to Use this Skill\" unless they also carry an NVIDIA, GPU,\naccelerated-computing, or distinctive NVIDIA workflow signal.",{"type":41,"tag":56,"props":170,"children":171},{},[172,174,181],{"type":46,"value":173},"Recommending a skill is always allowed once the request is relevant. Installing\nor modifying skills is not: never run an install (e.g. ",{"type":41,"tag":175,"props":176,"children":178},"code",{"className":177},[],[179],{"type":46,"value":180},"npx skills add",{"type":46,"value":182},") or\nchange agent capabilities without explicit user approval. A catalog match is\nonly a recommendation until the user confirms.",{"type":41,"tag":49,"props":184,"children":186},{"id":185},"when-not-to-use-this-skill",[187],{"type":46,"value":188},"When Not to Use this Skill",{"type":41,"tag":56,"props":190,"children":191},{},[192],{"type":46,"value":193},"Stay quiet when the request is generic:",{"type":41,"tag":83,"props":195,"children":196},{},[197,202,207,212,217],{"type":41,"tag":87,"props":198,"children":199},{},[200],{"type":46,"value":201},"\"route\" means an HTTP route, Express route, file route, or request routing.",{"type":41,"tag":87,"props":203,"children":204},{},[205],{"type":46,"value":206},"\"optimize\" means ordinary web performance, CSS, bundle size, SQL tuning, or\ngeneric code cleanup.",{"type":41,"tag":87,"props":208,"children":209},{},[210],{"type":46,"value":211},"\"deploy\" means generic Kubernetes, cloud, CI\u002FCD, or web hosting.",{"type":41,"tag":87,"props":213,"children":214},{},[215],{"type":46,"value":216},"\"AI\", \"data science\", or \"infrastructure\" appears without NVIDIA, GPU,\naccelerated-computing, or one of the distinctive intent signals above.",{"type":41,"tag":87,"props":218,"children":219},{},[220],{"type":46,"value":221},"\"video\" means ordinary trimming, captions, export, or social-media editing.",{"type":41,"tag":56,"props":223,"children":224},{},[225],{"type":46,"value":226},"If relevance is uncertain, do not interrupt the user's main task. Mention the\nNVIDIA catalog only as an optional aside after answering, or ask one concise\nclarifying question if the choice materially changes the work.",{"type":41,"tag":49,"props":228,"children":230},{"id":229},"instructions",[231],{"type":46,"value":232},"Instructions",{"type":41,"tag":56,"props":234,"children":235},{},[236],{"type":46,"value":237},"How to Help Users Find Skills - a Discovery Workflow",{"type":41,"tag":56,"props":239,"children":240},{},[241],{"type":46,"value":242},"This skill's first job is skill discovery. For NVIDIA-related requests, do a\ncatalog check before using general web search, NVIDIA product docs, or general\nproduct knowledge as the main answer. Product documentation can help after the\ncatalog check, but it is not a substitute for checking the NVIDIA skills catalog.",{"type":41,"tag":56,"props":244,"children":245},{},[246],{"type":46,"value":247},"Catalog check means one of:",{"type":41,"tag":83,"props":249,"children":250},{},[251,260,269,278],{"type":41,"tag":87,"props":252,"children":253},{},[254],{"type":41,"tag":175,"props":255,"children":257},{"className":256},[],[258],{"type":46,"value":259},"npx skills add nvidia\u002Fskills --list",{"type":41,"tag":87,"props":261,"children":262},{},[263],{"type":41,"tag":175,"props":264,"children":266},{"className":265},[],[267],{"type":46,"value":268},"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Fskills\u002Ftree\u002Fmain\u002Fskills",{"type":41,"tag":87,"props":270,"children":271},{},[272],{"type":41,"tag":175,"props":273,"children":275},{"className":274},[],[276],{"type":46,"value":277},"https:\u002F\u002Fbuild.nvidia.com\u002Fskills",{"type":41,"tag":87,"props":279,"children":280},{},[281],{"type":41,"tag":175,"props":282,"children":284},{"className":283},[],[285],{"type":46,"value":286},"https:\u002F\u002Fraw.githubusercontent.com\u002FNVIDIA\u002Fskills\u002Fmain\u002Fskills.sh.json",{"type":41,"tag":288,"props":289,"children":290},"ol",{},[291,296],{"type":41,"tag":87,"props":292,"children":293},{},[294],{"type":46,"value":295},"Check whether the relevant NVIDIA skill is already installed or already in context. If it is, hand off to that skill instead of recommending install.",{"type":41,"tag":87,"props":297,"children":298},{},[299],{"type":46,"value":300},"Query the live catalog before naming a specific install target or giving the\nmain product answer. When shell access is available, attempt this command first:",{"type":41,"tag":302,"props":303,"children":308},"pre",{"className":304,"code":305,"language":306,"meta":307,"style":307},"language-bash shiki shiki-themes material-theme-lighter material-theme material-theme-palenight","npx skills add nvidia\u002Fskills --list\n","bash","",[309],{"type":41,"tag":175,"props":310,"children":311},{"__ignoreMap":307},[312],{"type":41,"tag":313,"props":314,"children":317},"span",{"class":315,"line":316},"line",1,[318,324,330,335,340],{"type":41,"tag":313,"props":319,"children":321},{"style":320},"--shiki-light:#E2931D;--shiki-default:#FFCB6B;--shiki-dark:#FFCB6B",[322],{"type":46,"value":323},"npx",{"type":41,"tag":313,"props":325,"children":327},{"style":326},"--shiki-light:#91B859;--shiki-default:#C3E88D;--shiki-dark:#C3E88D",[328],{"type":46,"value":329}," skills",{"type":41,"tag":313,"props":331,"children":332},{"style":326},[333],{"type":46,"value":334}," add",{"type":41,"tag":313,"props":336,"children":337},{"style":326},[338],{"type":46,"value":339}," nvidia\u002Fskills",{"type":41,"tag":313,"props":341,"children":342},{"style":326},[343],{"type":46,"value":344}," --list\n",{"type":41,"tag":56,"props":346,"children":347},{},[348],{"type":46,"value":349},"Use the fallback catalog sources only if the CLI is unavailable, blocked, or fails:",{"type":41,"tag":83,"props":351,"children":352},{},[353,362,370],{"type":41,"tag":87,"props":354,"children":355},{},[356],{"type":41,"tag":146,"props":357,"children":360},{"href":268,"rel":358},[359],"nofollow",[361],{"type":46,"value":268},{"type":41,"tag":87,"props":363,"children":364},{},[365],{"type":41,"tag":146,"props":366,"children":368},{"href":277,"rel":367},[359],[369],{"type":46,"value":277},{"type":41,"tag":87,"props":371,"children":372},{},[373],{"type":41,"tag":146,"props":374,"children":376},{"href":286,"rel":375},[359],[377],{"type":46,"value":286},{"type":41,"tag":56,"props":379,"children":380},{},[381],{"type":46,"value":382},"Do not count a general web search, developer.nvidia.com product documentation,\nor docs.nvidia.com product documentation as the catalog check.",{"type":41,"tag":288,"props":384,"children":386},{"start":385},3,[387,392,397],{"type":41,"tag":87,"props":388,"children":389},{},[390],{"type":46,"value":391},"Match the request against current skill names, descriptions, product groups, and skill cards. Prefer NVIDIA-verified catalog entries over memory.",{"type":41,"tag":87,"props":393,"children":394},{},[395],{"type":46,"value":396},"If a current catalog skill strongly matches, recommend it before continuing\nwith general product guidance. Recommend at most three skills, ordered by\nconfidence. For each, include: skill name, why it fits, install command, and\nfirst useful prompt.",{"type":41,"tag":87,"props":398,"children":399},{},[400,402,407],{"type":46,"value":401},"Ask before installing. Do not run ",{"type":41,"tag":175,"props":403,"children":405},{"className":404},[],[406],{"type":46,"value":180},{"type":46,"value":408}," unless the user approves.",{"type":41,"tag":49,"props":410,"children":412},{"id":411},"recommendation-format",[413],{"type":46,"value":414},"Recommendation Format",{"type":41,"tag":56,"props":416,"children":417},{},[418],{"type":46,"value":419},"For a strong match that is not installed:",{"type":41,"tag":302,"props":421,"children":425},{"className":422,"code":424,"language":46,"meta":307},[423],"language-text","The NVIDIA \u003Cproduct or skill family> skill could help with this. Would you like me to install \u003Cskill-name>?\n",[426],{"type":41,"tag":175,"props":427,"children":428},{"__ignoreMap":307},[429],{"type":46,"value":424},{"type":41,"tag":56,"props":431,"children":432},{},[433],{"type":46,"value":434},"Use the active agent target when known:",{"type":41,"tag":302,"props":436,"children":438},{"className":304,"code":437,"language":306,"meta":307,"style":307},"npx skills add nvidia\u002Fskills --skill \u003Cskill-name> --agent codex --global --yes\nnpx skills add nvidia\u002Fskills --skill \u003Cskill-name> --agent claude-code --global --yes\n",[439],{"type":41,"tag":175,"props":440,"children":441},{"__ignoreMap":307},[442,508],{"type":41,"tag":313,"props":443,"children":444},{"class":315,"line":316},[445,449,453,457,461,466,472,477,483,488,493,498,503],{"type":41,"tag":313,"props":446,"children":447},{"style":320},[448],{"type":46,"value":323},{"type":41,"tag":313,"props":450,"children":451},{"style":326},[452],{"type":46,"value":329},{"type":41,"tag":313,"props":454,"children":455},{"style":326},[456],{"type":46,"value":334},{"type":41,"tag":313,"props":458,"children":459},{"style":326},[460],{"type":46,"value":339},{"type":41,"tag":313,"props":462,"children":463},{"style":326},[464],{"type":46,"value":465}," --skill",{"type":41,"tag":313,"props":467,"children":469},{"style":468},"--shiki-light:#39ADB5;--shiki-default:#89DDFF;--shiki-dark:#89DDFF",[470],{"type":46,"value":471}," \u003C",{"type":41,"tag":313,"props":473,"children":474},{"style":326},[475],{"type":46,"value":476},"skill-nam",{"type":41,"tag":313,"props":478,"children":480},{"style":479},"--shiki-light:#90A4AE;--shiki-default:#EEFFFF;--shiki-dark:#BABED8",[481],{"type":46,"value":482},"e",{"type":41,"tag":313,"props":484,"children":485},{"style":468},[486],{"type":46,"value":487},">",{"type":41,"tag":313,"props":489,"children":490},{"style":326},[491],{"type":46,"value":492}," --agent",{"type":41,"tag":313,"props":494,"children":495},{"style":326},[496],{"type":46,"value":497}," codex",{"type":41,"tag":313,"props":499,"children":500},{"style":326},[501],{"type":46,"value":502}," --global",{"type":41,"tag":313,"props":504,"children":505},{"style":326},[506],{"type":46,"value":507}," --yes\n",{"type":41,"tag":313,"props":509,"children":511},{"class":315,"line":510},2,[512,516,520,524,528,532,536,540,544,548,552,557,561],{"type":41,"tag":313,"props":513,"children":514},{"style":320},[515],{"type":46,"value":323},{"type":41,"tag":313,"props":517,"children":518},{"style":326},[519],{"type":46,"value":329},{"type":41,"tag":313,"props":521,"children":522},{"style":326},[523],{"type":46,"value":334},{"type":41,"tag":313,"props":525,"children":526},{"style":326},[527],{"type":46,"value":339},{"type":41,"tag":313,"props":529,"children":530},{"style":326},[531],{"type":46,"value":465},{"type":41,"tag":313,"props":533,"children":534},{"style":468},[535],{"type":46,"value":471},{"type":41,"tag":313,"props":537,"children":538},{"style":326},[539],{"type":46,"value":476},{"type":41,"tag":313,"props":541,"children":542},{"style":479},[543],{"type":46,"value":482},{"type":41,"tag":313,"props":545,"children":546},{"style":468},[547],{"type":46,"value":487},{"type":41,"tag":313,"props":549,"children":550},{"style":326},[551],{"type":46,"value":492},{"type":41,"tag":313,"props":553,"children":554},{"style":326},[555],{"type":46,"value":556}," claude-code",{"type":41,"tag":313,"props":558,"children":559},{"style":326},[560],{"type":46,"value":502},{"type":41,"tag":313,"props":562,"children":563},{"style":326},[564],{"type":46,"value":507},{"type":41,"tag":56,"props":566,"children":567},{},[568,570,576],{"type":46,"value":569},"If the agent target is unknown, omit ",{"type":41,"tag":175,"props":571,"children":573},{"className":572},[],[574],{"type":46,"value":575},"--agent",{"type":46,"value":577}," and let the CLI prompt:",{"type":41,"tag":302,"props":579,"children":581},{"className":304,"code":580,"language":306,"meta":307,"style":307},"npx skills add nvidia\u002Fskills --skill \u003Cskill-name> --global --yes\n",[582],{"type":41,"tag":175,"props":583,"children":584},{"__ignoreMap":307},[585],{"type":41,"tag":313,"props":586,"children":587},{"class":315,"line":316},[588,592,596,600,604,608,612,616,620,624,628],{"type":41,"tag":313,"props":589,"children":590},{"style":320},[591],{"type":46,"value":323},{"type":41,"tag":313,"props":593,"children":594},{"style":326},[595],{"type":46,"value":329},{"type":41,"tag":313,"props":597,"children":598},{"style":326},[599],{"type":46,"value":334},{"type":41,"tag":313,"props":601,"children":602},{"style":326},[603],{"type":46,"value":339},{"type":41,"tag":313,"props":605,"children":606},{"style":326},[607],{"type":46,"value":465},{"type":41,"tag":313,"props":609,"children":610},{"style":468},[611],{"type":46,"value":471},{"type":41,"tag":313,"props":613,"children":614},{"style":326},[615],{"type":46,"value":476},{"type":41,"tag":313,"props":617,"children":618},{"style":479},[619],{"type":46,"value":482},{"type":41,"tag":313,"props":621,"children":622},{"style":468},[623],{"type":46,"value":487},{"type":41,"tag":313,"props":625,"children":626},{"style":326},[627],{"type":46,"value":502},{"type":41,"tag":313,"props":629,"children":630},{"style":326},[631],{"type":46,"value":507},{"type":41,"tag":56,"props":633,"children":634},{},[635],{"type":46,"value":636},"After install, tell the user to restart or reload the agent if their client\ndoes not pick up newly installed skills immediately.",{"type":41,"tag":49,"props":638,"children":640},{"id":639},"confidence-rules",[641],{"type":46,"value":642},"Confidence Rules",{"type":41,"tag":83,"props":644,"children":645},{},[646,651,656],{"type":41,"tag":87,"props":647,"children":648},{},[649],{"type":46,"value":650},"High confidence: product name or distinctive intent maps cleanly to one\ncurrent catalog skill or skill family.",{"type":41,"tag":87,"props":652,"children":653},{},[654],{"type":46,"value":655},"Medium confidence: taxonomy lane matches but several skills may apply. Offer\na short shortlist and ask which direction matches the user's task.",{"type":41,"tag":87,"props":657,"children":658},{},[659],{"type":46,"value":660},"Low confidence: only generic wording matches. Do the user's task without\nrecommending an NVIDIA skill.",{"type":41,"tag":56,"props":662,"children":663},{},[664,666,672],{"type":46,"value":665},"Do not fabricate catalog entries or guess at skill names. If catalog lookup\nfails, say the catalog could not be checked and do not name a concrete slug or\nemit an ",{"type":41,"tag":175,"props":667,"children":669},{"className":668},[],[670],{"type":46,"value":671},"npx skills add ... --skill \u003Cname>",{"type":46,"value":673}," install command. Offer to share the\ncatalog URL, retry lookup, or continue with general help.",{"type":41,"tag":49,"props":675,"children":677},{"id":676},"when-no-skills-are-found",[678],{"type":46,"value":679},"When No Skills are Found",{"type":41,"tag":56,"props":681,"children":682},{},[683],{"type":46,"value":684},"Say that no strong NVIDIA catalog match was found, then either:",{"type":41,"tag":288,"props":686,"children":687},{},[688,699,704],{"type":41,"tag":87,"props":689,"children":690},{},[691,693],{"type":46,"value":692},"search the broader ecosystem with npx skills find ",{"type":41,"tag":694,"props":695,"children":696},"query",{},[697],{"type":46,"value":698},", or",{"type":41,"tag":87,"props":700,"children":701},{},[702],{"type":46,"value":703},"offer to help directly without a skill, or",{"type":41,"tag":87,"props":705,"children":706},{},[707],{"type":46,"value":708},"suggest creating a new skill if the workflow is recurring.",{"type":41,"tag":710,"props":711,"children":712},"style",{},[713],{"type":46,"value":714},"html .light .shiki span {color: var(--shiki-light);background: var(--shiki-light-bg);font-style: var(--shiki-light-font-style);font-weight: var(--shiki-light-font-weight);text-decoration: var(--shiki-light-text-decoration);}html.light .shiki span {color: var(--shiki-light);background: var(--shiki-light-bg);font-style: var(--shiki-light-font-style);font-weight: var(--shiki-light-font-weight);text-decoration: var(--shiki-light-text-decoration);}html .default .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}html.dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}",{"items":716,"total":875},[717,735,753,764,776,790,803,817,830,841,855,864],{"slug":718,"name":718,"fn":719,"description":720,"org":721,"tags":722,"stars":732,"repoUrl":733,"updatedAt":734},"nemoclaw-user-guide","retrieve NemoClaw documentation and configuration","Guides human users' AI agents to the NemoClaw docs MCP server and canonical Fern documentation in Markdown form. Use when users ask how to install, configure, operate, troubleshoot, secure, or learn NemoClaw with an AI coding assistant. Trigger keywords - nemoclaw docs, use nemoclaw with ai agent, nemoclaw mcp docs, nemoclaw install help, nemoclaw quickstart, nemoclaw markdown docs, llms.txt, agent skills.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[723,726,729],{"name":724,"slug":725,"type":15},"Documentation","documentation",{"name":727,"slug":728,"type":15},"MCP","mcp",{"name":730,"slug":731,"type":15},"Search","search",21777,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FNemoClaw","2026-07-20T06:00:01.461044",{"slug":736,"name":736,"fn":737,"description":738,"org":739,"tags":740,"stars":750,"repoUrl":751,"updatedAt":752},"mcore-build-and-dependency","manage Megatron-LM development environments","Container-based dev environment setup and dependency management for Megatron-LM. Covers acquiring and launching the CI container, uv package management, and updating uv.lock.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[741,744,747],{"name":742,"slug":743,"type":15},"Containers","containers",{"name":745,"slug":746,"type":15},"Deployment","deployment",{"name":748,"slug":749,"type":15},"Python","python",17049,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FMegatron-LM","2026-07-27T06:06:11.249662",{"slug":754,"name":754,"fn":755,"description":756,"org":757,"tags":758,"stars":750,"repoUrl":751,"updatedAt":763},"mcore-bump-base-image","update NVIDIA PyTorch base images","Bump the NVIDIA PyTorch base image (`nvcr.io\u002Fnvidia\u002Fpytorch:YY.MM-py3`) used by Megatron-LM CI. Covers the two pin sites (GitHub CI in `docker\u002F.ngc_version.dev` and GitLab CI in `.gitlab\u002Fstages\u002F01.build.yml`), the post-bump CI loop (re-run functional tests, refresh golden values, mark broken tests), and the gotchas that bit PRs",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[759,762],{"name":760,"slug":761,"type":15},"CI\u002FCD","ci-cd",{"name":745,"slug":746,"type":15},"2026-07-14T05:25:59.97109",{"slug":765,"name":765,"fn":766,"description":767,"org":768,"tags":769,"stars":750,"repoUrl":751,"updatedAt":775},"mcore-cicd","manage CI\u002FCD pipelines for Megatron-LM","CI\u002FCD reference for Megatron-LM. Covers CI pipeline structure, PR scope labels, triggering internal GitLab CI (which force-pushes the current branch to a pull-request\u002FBRANCH ref — always dry-run and verify the destination first; never run against shared or protected branches), and CI failure investigation.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[770,771,772],{"name":760,"slug":761,"type":15},{"name":745,"slug":746,"type":15},{"name":773,"slug":774,"type":15},"GitHub","github","2026-07-27T06:06:12.278222",{"slug":777,"name":777,"fn":778,"description":779,"org":780,"tags":781,"stars":750,"repoUrl":751,"updatedAt":789},"mcore-create-issue","investigate CI failures and create issues","Investigate a failing GitHub Actions run or job and create a GitHub issue for the failure.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[782,785,786],{"name":783,"slug":784,"type":15},"Debugging","debugging",{"name":773,"slug":774,"type":15},{"name":787,"slug":788,"type":15},"Triage","triage","2026-07-14T05:25:57.442089",{"slug":791,"name":791,"fn":792,"description":793,"org":794,"tags":795,"stars":750,"repoUrl":751,"updatedAt":802},"mcore-linting-and-formatting","lint and format Megatron-LM code","Linting and formatting for Megatron-LM. Covers running autoformat.sh, tools (ruff, black, isort, pylint, mypy), and code style rules.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[796,799],{"name":797,"slug":798,"type":15},"Best Practices","best-practices",{"name":800,"slug":801,"type":15},"Code Analysis","code-analysis","2026-07-14T05:25:56.18433",{"slug":804,"name":804,"fn":805,"description":806,"org":807,"tags":808,"stars":750,"repoUrl":751,"updatedAt":816},"mcore-migrate-gpt-to-hybrid","migrate Megatron-LM models to HybridModel","Migration guide for moving Megatron Core GPTModel checkpoints, model providers, training commands, and layer mappings to HybridModel.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[809,812,815],{"name":810,"slug":811,"type":15},"Machine Learning","machine-learning",{"name":813,"slug":814,"type":15},"Migration","migration",{"name":9,"slug":8,"type":15},"2026-07-17T06:07:11.777011",{"slug":818,"name":818,"fn":819,"description":820,"org":821,"tags":822,"stars":750,"repoUrl":751,"updatedAt":829},"mcore-onboard-gb200-1node-tests","onboard functional tests for GB200","Onboard 1-node GitHub MR functional tests for GB200 from existing mr-scoped 2-node tests.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[823,826],{"name":824,"slug":825,"type":15},"QA","qa",{"name":827,"slug":828,"type":15},"Testing","testing","2026-07-14T05:25:53.673039",{"slug":831,"name":831,"fn":832,"description":833,"org":834,"tags":835,"stars":750,"repoUrl":751,"updatedAt":840},"mcore-run-on-slurm","launch distributed training jobs on SLURM","How to launch distributed Megatron-LM training jobs on a SLURM cluster. Covers a minimal sbatch skeleton, environment-variable setup for torch.distributed.run, CUDA_DEVICE_MAX_CONNECTIONS rules across hardware and parallelism modes, container conventions, monitoring, and per-rank failure diagnosis.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[836,837],{"name":745,"slug":746,"type":15},{"name":838,"slug":839,"type":15},"Infrastructure","infrastructure","2026-07-14T05:25:49.362534",{"slug":842,"name":842,"fn":843,"description":844,"org":845,"tags":846,"stars":750,"repoUrl":751,"updatedAt":854},"mcore-split-pr","split pull requests to reduce review load","Split a PR into multiple PRs to reduce the number of required CODEOWNERS reviewer groups.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[847,850,851],{"name":848,"slug":849,"type":15},"Code Review","code-review",{"name":773,"slug":774,"type":15},{"name":852,"slug":853,"type":15},"Pull Requests","pull-requests","2026-07-14T05:26:01.226578",{"slug":856,"name":856,"fn":857,"description":858,"org":859,"tags":860,"stars":750,"repoUrl":751,"updatedAt":863},"mcore-testing","run and manage Megatron-LM tests","Test system for Megatron-LM. Covers test layout, recipe YAML structure, adding and running unit and functional tests, golden values, marker filters, and CI parity.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[861,862],{"name":824,"slug":825,"type":15},{"name":827,"slug":828,"type":15},"2026-07-14T05:25:54.928983",{"slug":865,"name":865,"fn":866,"description":867,"org":868,"tags":869,"stars":750,"repoUrl":751,"updatedAt":874},"nightly-sync","manage nightly main-to-dev sync workflows","Domain knowledge for the nightly main-to-dev sync workflow. Covers merge strategy, CI architecture, failure investigation, and known issues.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[870,873],{"name":871,"slug":872,"type":15},"Automation","automation",{"name":760,"slug":761,"type":15},"2026-07-30T05:29:03.275638",496,{"items":877,"total":973},[878,895,905,919,929,944,959],{"slug":879,"name":879,"fn":880,"description":881,"org":882,"tags":883,"stars":17,"repoUrl":18,"updatedAt":894},"accelerated-computing-cudf","accelerate data processing with cuDF","Official NVIDIA-authored guidance for NVIDIA cuDF GPU DataFrames, pandas acceleration, dask-cuDF, ETL, joins, groupby, CSV\u002FParquet I\u002FO, nullable semantics, and multi-GPU DataFrame workloads.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[884,887,890,891],{"name":885,"slug":886,"type":15},"Data Analysis","data-analysis",{"name":888,"slug":889,"type":15},"Data Engineering","data-engineering",{"name":9,"slug":8,"type":15},{"name":892,"slug":893,"type":15},"Performance","performance","2026-07-14T05:28:43.176466",{"slug":896,"name":896,"fn":897,"description":898,"org":899,"tags":900,"stars":17,"repoUrl":18,"updatedAt":904},"aiq-deploy","deploy and manage NVIDIA AI-Q infrastructure","Use when asked to install, deploy, run, validate, troubleshoot, or stop NVIDIA AI-Q Blueprint infrastructure.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[901,902,903],{"name":745,"slug":746,"type":15},{"name":838,"slug":839,"type":15},{"name":9,"slug":8,"type":15},"2026-07-14T05:29:06.667109",{"slug":906,"name":906,"fn":907,"description":908,"org":909,"tags":910,"stars":17,"repoUrl":18,"updatedAt":918},"aiq-research","conduct deep research with AI-Q","Use when asked to run deep research or AI-Q research through a reachable NVIDIA AI-Q Blueprint backend.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[911,914,915],{"name":912,"slug":913,"type":15},"Agents","agents",{"name":9,"slug":8,"type":15},{"name":916,"slug":917,"type":15},"Research","research","2026-07-14T05:28:06.816956",{"slug":920,"name":920,"fn":921,"description":922,"org":923,"tags":924,"stars":17,"repoUrl":18,"updatedAt":928},"amc-run-sample-calibration","run AMC sample dataset calibration","Run end-to-end calibration on the shipped sample dataset (sdg_08_2_sample_data_010926.zip) against a running AMC microservice. Use when user says 'test sample dataset', 'run sample calibration', 'verify AMC install', or 'launch and test'.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[925,926,927],{"name":885,"slug":886,"type":15},{"name":9,"slug":8,"type":15},{"name":827,"slug":828,"type":15},"2026-07-17T05:29:03.913266",{"slug":930,"name":930,"fn":931,"description":932,"org":933,"tags":934,"stars":17,"repoUrl":18,"updatedAt":943},"amc-run-video-calibration","calibrate video datasets with AutoMagicCalib","Calibrate a new dataset from pre-recorded video files via the AutoMagicCalib REST API. Use when user has local MP4s and says 'calibrate my videos', 'run AMC on these videos', or similar. For RTSP\u002Flive streams, use amc-run-rtsp-calibration instead.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[935,936,939,940],{"name":871,"slug":872,"type":15},{"name":937,"slug":938,"type":15},"Imaging","imaging",{"name":9,"slug":8,"type":15},{"name":941,"slug":942,"type":15},"Video","video","2026-07-17T05:28:53.905004",{"slug":945,"name":945,"fn":946,"description":947,"org":948,"tags":949,"stars":17,"repoUrl":18,"updatedAt":958},"amc-setup-calibration-stack","deploy AutoMagicCalib microservice with Docker","Launch AutoMagicCalib microservice and web UI from NGC release images via Docker Compose. Use when user says 'deploy auto calibration', 'launch auto calibration', 'launch AMC', 'start MS+UI', or 'set up auto-magic-calib'. Requires NGC API key.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[950,951,954,955],{"name":745,"slug":746,"type":15},{"name":952,"slug":953,"type":15},"Docker","docker",{"name":9,"slug":8,"type":15},{"name":956,"slug":957,"type":15},"Operations","operations","2026-07-17T05:28:56.913999",{"slug":960,"name":960,"fn":961,"description":962,"org":963,"tags":964,"stars":17,"repoUrl":18,"updatedAt":972},"cudaq-guide","develop quantum applications with CUDA-Q","CUDA-Q onboarding guide for installation, test programs, GPU simulation, QPU hardware, and quantum applications.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[965,966,969],{"name":9,"slug":8,"type":15},{"name":967,"slug":968,"type":15},"Quantum Computing","quantum-computing",{"name":970,"slug":971,"type":15},"Simulation","simulation","2026-07-14T05:26:58.898253",305]