[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill-nvidia-dgx-station-mig":3,"mdc--iavbyl-key":31,"related-repo-nvidia-dgx-station-mig":247,"related-org-nvidia-dgx-station-mig":353},{"slug":4,"name":4,"fn":5,"description":6,"org":7,"tags":11,"stars":20,"repoUrl":21,"updatedAt":22,"license":23,"forks":24,"topics":25,"repo":26,"sourceUrl":29,"mdContent":30},"dgx-station-mig","manage NVIDIA DGX Station MIG profiles","Inspect NVIDIA DGX Station GB300 MIG state and installed-driver profiles, create a digest-bound layout plan, disclose disruption and restoration, and apply an approved still-valid plan. Use when the user asks to enable, disable, partition, reconfigure, inspect, or troubleshoot MIG instances or needs MIG UUIDs. Never assume static profile IDs or terminate GPU clients.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},"nvidia","NVIDIA","https:\u002F\u002Fpexgzepcugksgbtrxkhf.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Forg-logos\u002Fnvidia.png",[12,16,17],{"name":13,"slug":14,"type":15},"Operations","operations","tag",{"name":9,"slug":8,"type":15},{"name":18,"slug":19,"type":15},"Engineering","engineering",1144,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Fdgx-spark-playbooks","2026-08-05T05:58:16.344974",null,249,[],{"repoUrl":21,"stars":20,"forks":24,"topics":27,"description":28},[],"Collection of step-by-step playbooks for setting up AI\u002FML workloads on NVIDIA DGX Spark devices with Blackwell architecture.","https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Fdgx-spark-playbooks\u002Ftree\u002FHEAD\u002Fnvidia\u002Fstation-ai-skills\u002Fassets\u002Fskills\u002Fdgx-station-mig","---\nname: dgx-station-mig\ndescription: Inspect NVIDIA DGX Station GB300 MIG state and installed-driver profiles, create a digest-bound layout plan, disclose disruption and restoration, and apply an approved still-valid plan. Use when the user asks to enable, disable, partition, reconfigure, inspect, or troubleshoot MIG instances or needs MIG UUIDs. Never assume static profile IDs or terminate GPU clients.\n---\n\n# DGX Station MIG\n\nTreat inspection and planning as read-only. Treat every apply as disruptive.\n\n## Workflow\n\n1. Run `scripts\u002Fdgx-assist system inspect --json`.\n2. Search the playbooks for the user's MIG concern and cite relevant results.\n3. Read `compatibility.capabilities.mig_inspection`. If it is false, stop and explain that the detected release has no MIG inspection profile.\n4. Run `mig inspect` and `mig profiles`. Use only profiles reported by the installed driver. This read-only step is available on the recognized Software 1.0 and Software 2.0 profiles.\n5. Before planning, require `compatibility.capabilities.mig_mutation` to be `true`. If it is false, report the observed state but do not suggest a layout or work around the restriction.\n6. Require the desired exact layout. Do not choose a layout from model names or nominal memory sums.\n7. Run `mig plan --layout \"\u003Clayout>\"`.\n8. Present current and proposed state, exact commands, active clients, privilege, workload impact, reset\u002Freboot possibility, and restoration commands.\n9. If any client is active, stop. Tell the user dgx-assist will not terminate it.\n10. Obtain explicit confirmation immediately before mutation.\n11. Run `mig apply --plan-id \"\u003Cid>\" --yes`.\n12. Report every verified MIG UUID. On partial failure, stop and present only the recorded restoration plan.\n\n## Safety requirements\n\n- Never use a hardcoded GPU index, profile table, or assumed placement.\n- Never destroy instances, reset a GPU, reboot, or change MIG mode outside an approved plan.\n- Never kill a GPU process or take over a service.\n- Invalidate the plan when clients, mode, profiles, release, driver, or instances change.\n- Do not improvise further mutations after partial failure.\n- Do not treat Fabric Manager as a normal Station prerequisite.\n- Treat `--yes` only as approval already obtained.\n\nRead [references\u002Fworkflow.md](references\u002Fworkflow.md) before planning or applying a layout.\n",{"data":32,"body":33},{"name":4,"description":6},{"type":34,"children":35},"root",[36,44,50,57,182,188,235],{"type":37,"tag":38,"props":39,"children":40},"element","h1",{"id":4},[41],{"type":42,"value":43},"text","DGX Station MIG",{"type":37,"tag":45,"props":46,"children":47},"p",{},[48],{"type":42,"value":49},"Treat inspection and planning as read-only. Treat every apply as disruptive.",{"type":37,"tag":51,"props":52,"children":54},"h2",{"id":53},"workflow",[55],{"type":42,"value":56},"Workflow",{"type":37,"tag":58,"props":59,"children":60},"ol",{},[61,76,81,94,114,135,140,151,156,161,166,177],{"type":37,"tag":62,"props":63,"children":64},"li",{},[65,67,74],{"type":42,"value":66},"Run ",{"type":37,"tag":68,"props":69,"children":71},"code",{"className":70},[],[72],{"type":42,"value":73},"scripts\u002Fdgx-assist system inspect --json",{"type":42,"value":75},".",{"type":37,"tag":62,"props":77,"children":78},{},[79],{"type":42,"value":80},"Search the playbooks for the user's MIG concern and cite relevant results.",{"type":37,"tag":62,"props":82,"children":83},{},[84,86,92],{"type":42,"value":85},"Read ",{"type":37,"tag":68,"props":87,"children":89},{"className":88},[],[90],{"type":42,"value":91},"compatibility.capabilities.mig_inspection",{"type":42,"value":93},". If it is false, stop and explain that the detected release has no MIG inspection profile.",{"type":37,"tag":62,"props":95,"children":96},{},[97,98,104,106,112],{"type":42,"value":66},{"type":37,"tag":68,"props":99,"children":101},{"className":100},[],[102],{"type":42,"value":103},"mig inspect",{"type":42,"value":105}," and ",{"type":37,"tag":68,"props":107,"children":109},{"className":108},[],[110],{"type":42,"value":111},"mig profiles",{"type":42,"value":113},". Use only profiles reported by the installed driver. This read-only step is available on the recognized Software 1.0 and Software 2.0 profiles.",{"type":37,"tag":62,"props":115,"children":116},{},[117,119,125,127,133],{"type":42,"value":118},"Before planning, require ",{"type":37,"tag":68,"props":120,"children":122},{"className":121},[],[123],{"type":42,"value":124},"compatibility.capabilities.mig_mutation",{"type":42,"value":126}," to be ",{"type":37,"tag":68,"props":128,"children":130},{"className":129},[],[131],{"type":42,"value":132},"true",{"type":42,"value":134},". If it is false, report the observed state but do not suggest a layout or work around the restriction.",{"type":37,"tag":62,"props":136,"children":137},{},[138],{"type":42,"value":139},"Require the desired exact layout. Do not choose a layout from model names or nominal memory sums.",{"type":37,"tag":62,"props":141,"children":142},{},[143,144,150],{"type":42,"value":66},{"type":37,"tag":68,"props":145,"children":147},{"className":146},[],[148],{"type":42,"value":149},"mig plan --layout \"\u003Clayout>\"",{"type":42,"value":75},{"type":37,"tag":62,"props":152,"children":153},{},[154],{"type":42,"value":155},"Present current and proposed state, exact commands, active clients, privilege, workload impact, reset\u002Freboot possibility, and restoration commands.",{"type":37,"tag":62,"props":157,"children":158},{},[159],{"type":42,"value":160},"If any client is active, stop. Tell the user dgx-assist will not terminate it.",{"type":37,"tag":62,"props":162,"children":163},{},[164],{"type":42,"value":165},"Obtain explicit confirmation immediately before mutation.",{"type":37,"tag":62,"props":167,"children":168},{},[169,170,176],{"type":42,"value":66},{"type":37,"tag":68,"props":171,"children":173},{"className":172},[],[174],{"type":42,"value":175},"mig apply --plan-id \"\u003Cid>\" --yes",{"type":42,"value":75},{"type":37,"tag":62,"props":178,"children":179},{},[180],{"type":42,"value":181},"Report every verified MIG UUID. On partial failure, stop and present only the recorded restoration plan.",{"type":37,"tag":51,"props":183,"children":185},{"id":184},"safety-requirements",[186],{"type":42,"value":187},"Safety requirements",{"type":37,"tag":189,"props":190,"children":191},"ul",{},[192,197,202,207,212,217,222],{"type":37,"tag":62,"props":193,"children":194},{},[195],{"type":42,"value":196},"Never use a hardcoded GPU index, profile table, or assumed placement.",{"type":37,"tag":62,"props":198,"children":199},{},[200],{"type":42,"value":201},"Never destroy instances, reset a GPU, reboot, or change MIG mode outside an approved plan.",{"type":37,"tag":62,"props":203,"children":204},{},[205],{"type":42,"value":206},"Never kill a GPU process or take over a service.",{"type":37,"tag":62,"props":208,"children":209},{},[210],{"type":42,"value":211},"Invalidate the plan when clients, mode, profiles, release, driver, or instances change.",{"type":37,"tag":62,"props":213,"children":214},{},[215],{"type":42,"value":216},"Do not improvise further mutations after partial failure.",{"type":37,"tag":62,"props":218,"children":219},{},[220],{"type":42,"value":221},"Do not treat Fabric Manager as a normal Station prerequisite.",{"type":37,"tag":62,"props":223,"children":224},{},[225,227,233],{"type":42,"value":226},"Treat ",{"type":37,"tag":68,"props":228,"children":230},{"className":229},[],[231],{"type":42,"value":232},"--yes",{"type":42,"value":234}," only as approval already obtained.",{"type":37,"tag":45,"props":236,"children":237},{},[238,239,245],{"type":42,"value":85},{"type":37,"tag":240,"props":241,"children":243},"a",{"href":242},"references\u002Fworkflow.md",[244],{"type":42,"value":242},{"type":42,"value":246}," before planning or applying a layout.",{"items":248,"total":352},[249,269,282,297,312,323,340],{"slug":250,"name":250,"fn":251,"description":252,"org":253,"tags":254,"stars":20,"repoUrl":21,"updatedAt":268},"analysis-methods","write Python analysis code for FHIR data","Teaches the analyst agent how to write correct, robust Python analysis code for FHIR clinical data using pandas, matplotlib, and scipy.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[255,258,261,264,265],{"name":256,"slug":257,"type":15},"Data Analysis","data-analysis",{"name":259,"slug":260,"type":15},"FHIR","fhir",{"name":262,"slug":263,"type":15},"Healthcare","healthcare",{"name":9,"slug":8,"type":15},{"name":266,"slug":267,"type":15},"Python","python","2026-07-14T05:35:59.037962",{"slug":270,"name":270,"fn":271,"description":272,"org":273,"tags":274,"stars":20,"repoUrl":21,"updatedAt":281},"case-summary","summarize clinical patient cases from FHIR","Prepare a complete clinical case summary for a patient from FHIR endpoints. Use when asked to summarize a patient, compile a case, or prepare for tumor board.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[275,276,277,278],{"name":259,"slug":260,"type":15},{"name":262,"slug":263,"type":15},{"name":9,"slug":8,"type":15},{"name":279,"slug":280,"type":15},"Summarization","summarization","2026-07-14T05:35:52.790528",{"slug":283,"name":283,"fn":284,"description":285,"org":286,"tags":287,"stars":20,"repoUrl":21,"updatedAt":296},"clinical-delegation","delegate clinical tasks to specialist agents","How to delegate clinical tasks to specialist agents. Always use sub-agent runtime with explicit agentId — never ACP. Never call FHIR via web_fetch.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[288,291,292,295],{"name":289,"slug":290,"type":15},"Agents","agents",{"name":262,"slug":263,"type":15},{"name":293,"slug":294,"type":15},"Multi-Agent","multi-agent",{"name":9,"slug":8,"type":15},"2026-07-14T05:35:55.294972",{"slug":298,"name":298,"fn":299,"description":300,"org":301,"tags":302,"stars":20,"repoUrl":21,"updatedAt":311},"clinical-knowledge","provide clinical reference and regulatory context","Teaches agents clinical reference ranges, condition codes, quality measure definitions, drug classifications, and regulatory context so they can flag abnormal values and identify care gaps.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[303,306,307,308],{"name":304,"slug":305,"type":15},"Clinical Trials","clinical-trials",{"name":262,"slug":263,"type":15},{"name":9,"slug":8,"type":15},{"name":309,"slug":310,"type":15},"Regulatory Compliance","regulatory-compliance","2026-07-14T05:35:56.550833",{"slug":313,"name":313,"fn":314,"description":315,"org":316,"tags":317,"stars":20,"repoUrl":21,"updatedAt":322},"cohort-compare","analyze patient cohorts from FHIR endpoints","Analyze a cohort of patients from FHIR endpoints to find care gaps and patterns. Use when asked to compare patients, find quality gaps, or analyze a population.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[318,319,320,321],{"name":256,"slug":257,"type":15},{"name":259,"slug":260,"type":15},{"name":262,"slug":263,"type":15},{"name":9,"slug":8,"type":15},"2026-07-14T05:35:51.543095",{"slug":324,"name":324,"fn":325,"description":326,"org":327,"tags":328,"stars":20,"repoUrl":21,"updatedAt":339},"dgx-diagnose","diagnose NVIDIA DGX Station hardware issues","Diagnose common DGX Station GB300 issues — CUDA crashes, wrong-GPU targeting, vLLM\u002FSGLang container bugs, MIG state problems, NVLink\u002FFabric Manager errors, X\u002FVulkan failures, HuggingFace auth, and port conflicts. Use when the user reports a GPU error, inference server crash, MIG problem, or any unexplained DGX Station failure.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[329,332,335,336],{"name":330,"slug":331,"type":15},"AI Infrastructure","ai-infrastructure",{"name":333,"slug":334,"type":15},"Debugging","debugging",{"name":9,"slug":8,"type":15},{"name":337,"slug":338,"type":15},"Observability","observability","2026-07-14T05:31:04.085598",{"slug":341,"name":341,"fn":342,"description":343,"org":344,"tags":345,"stars":20,"repoUrl":21,"updatedAt":351},"dgx-station","manage NVIDIA DGX Station development","Inspect and guide NVIDIA DGX Station GB300 development using the local dgx-assist CLI and pinned NVIDIA playbooks. Use for general Station platform questions, Software 1.0 or 2.0 compatibility, GB300 or RTX GPU selection, UUID ordering, mixed ATS\u002FHMM coherency, CDMM, general containers, CDI, CUDA visibility, or vsloshd power-sloshing behavior. Do not use for vLLM or SGLang container selection or tuning, serving a named model, changing MIG, or troubleshooting a reported failure when the dedicated Station skill applies.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[346,349,350],{"name":347,"slug":348,"type":15},"CLI","cli",{"name":9,"slug":8,"type":15},{"name":13,"slug":14,"type":15},"2026-08-05T05:58:16.736667",15,{"items":354,"total":509},[355,373,389,400,412,424,437,451,464,475,489,498],{"slug":356,"name":356,"fn":357,"description":358,"org":359,"tags":360,"stars":370,"repoUrl":371,"updatedAt":372},"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},[361,364,367],{"name":362,"slug":363,"type":15},"Documentation","documentation",{"name":365,"slug":366,"type":15},"MCP","mcp",{"name":368,"slug":369,"type":15},"Search","search",21777,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FNemoClaw","2026-07-20T06:00:01.461044",{"slug":374,"name":374,"fn":375,"description":376,"org":377,"tags":378,"stars":386,"repoUrl":387,"updatedAt":388},"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},[379,382,385],{"name":380,"slug":381,"type":15},"Containers","containers",{"name":383,"slug":384,"type":15},"Deployment","deployment",{"name":266,"slug":267,"type":15},17049,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FMegatron-LM","2026-07-27T06:06:11.249662",{"slug":390,"name":390,"fn":391,"description":392,"org":393,"tags":394,"stars":386,"repoUrl":387,"updatedAt":399},"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},[395,398],{"name":396,"slug":397,"type":15},"CI\u002FCD","ci-cd",{"name":383,"slug":384,"type":15},"2026-07-14T05:25:59.97109",{"slug":401,"name":401,"fn":402,"description":403,"org":404,"tags":405,"stars":386,"repoUrl":387,"updatedAt":411},"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},[406,407,408],{"name":396,"slug":397,"type":15},{"name":383,"slug":384,"type":15},{"name":409,"slug":410,"type":15},"GitHub","github","2026-07-27T06:06:12.278222",{"slug":413,"name":413,"fn":414,"description":415,"org":416,"tags":417,"stars":386,"repoUrl":387,"updatedAt":423},"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},[418,419,420],{"name":333,"slug":334,"type":15},{"name":409,"slug":410,"type":15},{"name":421,"slug":422,"type":15},"Triage","triage","2026-07-14T05:25:57.442089",{"slug":425,"name":425,"fn":426,"description":427,"org":428,"tags":429,"stars":386,"repoUrl":387,"updatedAt":436},"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},[430,433],{"name":431,"slug":432,"type":15},"Best Practices","best-practices",{"name":434,"slug":435,"type":15},"Code Analysis","code-analysis","2026-07-14T05:25:56.18433",{"slug":438,"name":438,"fn":439,"description":440,"org":441,"tags":442,"stars":386,"repoUrl":387,"updatedAt":450},"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},[443,446,449],{"name":444,"slug":445,"type":15},"Machine Learning","machine-learning",{"name":447,"slug":448,"type":15},"Migration","migration",{"name":9,"slug":8,"type":15},"2026-07-17T06:07:11.777011",{"slug":452,"name":452,"fn":453,"description":454,"org":455,"tags":456,"stars":386,"repoUrl":387,"updatedAt":463},"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},[457,460],{"name":458,"slug":459,"type":15},"QA","qa",{"name":461,"slug":462,"type":15},"Testing","testing","2026-07-14T05:25:53.673039",{"slug":465,"name":465,"fn":466,"description":467,"org":468,"tags":469,"stars":386,"repoUrl":387,"updatedAt":474},"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},[470,471],{"name":383,"slug":384,"type":15},{"name":472,"slug":473,"type":15},"Infrastructure","infrastructure","2026-07-14T05:25:49.362534",{"slug":476,"name":476,"fn":477,"description":478,"org":479,"tags":480,"stars":386,"repoUrl":387,"updatedAt":488},"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},[481,484,485],{"name":482,"slug":483,"type":15},"Code Review","code-review",{"name":409,"slug":410,"type":15},{"name":486,"slug":487,"type":15},"Pull Requests","pull-requests","2026-07-14T05:26:01.226578",{"slug":490,"name":490,"fn":491,"description":492,"org":493,"tags":494,"stars":386,"repoUrl":387,"updatedAt":497},"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},[495,496],{"name":458,"slug":459,"type":15},{"name":461,"slug":462,"type":15},"2026-07-14T05:25:54.928983",{"slug":499,"name":499,"fn":500,"description":501,"org":502,"tags":503,"stars":386,"repoUrl":387,"updatedAt":508},"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},[504,507],{"name":505,"slug":506,"type":15},"Automation","automation",{"name":396,"slug":397,"type":15},"2026-07-30T05:29:03.275638",525]