[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill-nvidia-doca-caps":3,"mdc-g983pm-key":31,"related-repo-nvidia-doca-caps":1137,"related-org-nvidia-doca-caps":1243},{"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},"doca-caps","query DOCA device and host capabilities","Use this skill when the user wants to invoke the read-only doca_caps CLI to ask what DOCA sees on this host — listing DOCA devices and PCIe addresses, listing representor devices, asking which DOCA libraries are available on the current OS, checking per-device per-library capabilities, scoping output to a specific PCIe address, or capturing a side-effect-free capability snapshot for a debug session or install smoke-test. Trigger even when the user does not explicitly mention \"doca_caps\" or \"capabilities print tool\" — typical implicit phrasings include \"what does DOCA actually see on this box\", \"is my BlueField PF visible to DOCA\", \"is Flow available on my RHEL host\", \"enumerate VF representors for pf0\", \"doca_caps: command not found\", or \"empty output for RDMA, is the tool broken\". Refuse and route elsewhere for DOCA installation, library-internal capability matrices (Flow pipe creation, RDMA verbs features), streaming telemetry \u002F DTS, or modifying the shipped binary — those belong to other skills.\n",{"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},"CLI","cli","tag",{"name":9,"slug":8,"type":15},{"name":18,"slug":19,"type":15},"Engineering","engineering",2473,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Fskills","2026-07-20T06:24:01.551181","Apache-2.0",281,[],{"repoUrl":21,"stars":20,"forks":24,"topics":27,"description":28},[],"AI agent skills published by NVIDIA","https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Fskills\u002Ftree\u002FHEAD\u002Fskills\u002Fdoca-caps","---\nlicense: Apache-2.0\nname: doca-caps\ndescription: >\n  Use this skill when the user wants to invoke the read-only\n  doca_caps CLI to ask what DOCA sees on this host — listing\n  DOCA devices and PCIe addresses, listing representor devices,\n  asking which DOCA libraries are available on the current OS,\n  checking per-device per-library capabilities, scoping output\n  to a specific PCIe address, or capturing a side-effect-free\n  capability snapshot for a debug session or install\n  smoke-test. Trigger even when the user does not explicitly\n  mention \"doca_caps\" or \"capabilities print tool\" — typical\n  implicit phrasings include \"what does DOCA actually see on\n  this box\", \"is my BlueField PF visible to DOCA\", \"is Flow\n  available on my RHEL host\", \"enumerate VF representors for\n  pf0\", \"doca_caps: command not found\", or \"empty output for\n  RDMA, is the tool broken\". Refuse and route elsewhere for\n  DOCA installation, library-internal capability matrices\n  (Flow pipe creation, RDMA verbs features), streaming\n  telemetry \u002F DTS, or modifying the shipped binary — those\n  belong to other skills.\nmetadata:\n  kind: tool\ncompatibility: >\n  Requires DOCA SDK ≥ 2.6.0 installed at \u002Fopt\u002Fmellanox\u002Fdoca\n  on Linux (Ubuntu 22.04\u002F24.04 or RHEL\u002FSLES) with a BlueField\n  DPU or ConnectX NIC; runs identically on the host or on the\n  BlueField Arm side. Invokes\n  \u002Fopt\u002Fmellanox\u002Fdoca\u002Ftools\u002Fdoca_caps and reads\n  `pkg-config doca-common` to confirm the install.\n---\n\n# DOCA Capabilities Print Tool (`doca_caps`)\n\n**Where to start:** This is a tool skill for invoking `doca_caps`,\na side-effect-free CLI. Open [`TASKS.md`](TASKS.md) and start at\n[`## run`](TASKS.md#run) for the documented invocations, or\n[`## test`](TASKS.md#test) when using `doca_caps` as an install\nsmoke-test. Open [`CAPABILITIES.md`](CAPABILITIES.md) when the\nquestion is *what kinds of capability families `doca_caps` reports*.\nIf DOCA is not installed yet, route to\n[`doca-setup`](..\u002F..\u002Fdoca-setup\u002FSKILL.md) first.\n\n## Example questions this skill answers well\n\nThe CLASSES of `doca_caps` questions this skill is built to answer,\neach with one worked example. The class is the load-bearing piece;\nthe worked example is one instance.\n\n- **\"What DOCA devices does this host see?\"** — worked example: *\"is\n  my BlueField PF visible to DOCA\"*. Answered by the device\n  enumeration in\n  [`CAPABILITIES.md ## Capabilities and modes`](CAPABILITIES.md#capabilities-and-modes)\n  + the `--list-devs` invocation in\n  [`TASKS.md ## run`](TASKS.md#run).\n- **\"Which DOCA libraries are available on this OS?\"** — worked\n  example: *\"is Flow available on my RHEL host\"*. Answered by the\n  library-availability surface in\n  [`CAPABILITIES.md ## Capabilities and modes`](CAPABILITIES.md#capabilities-and-modes)\n  + the library-listing invocation in\n  [`TASKS.md ## run`](TASKS.md#run).\n- **\"Does this device support library X capability Y?\"** — worked\n  example: *\"does this device support Flow hairpin?\"*. Answered by\n  the per-device per-library capability matrix in\n  [`CAPABILITIES.md ## Capabilities and modes`](CAPABILITIES.md#capabilities-and-modes)\n  + the `--pci-addr`-scoped invocation in\n  [`TASKS.md ## run`](TASKS.md#run).\n- **\"What representors are visible to DOCA?\"** — worked example:\n  *\"enumerate VF representors for pf0\"*. Answered by the representor\n  enumeration in\n  [`CAPABILITIES.md ## Capabilities and modes`](CAPABILITIES.md#capabilities-and-modes)\n  + the `--list-rep-devs` invocation in\n  [`TASKS.md ## run`](TASKS.md#run).\n- **\"I want a snapshot of state to attach to my debug session.\"** —\n  worked example: *\"save device + library + capability output to a\n  file\"*. Answered by the snapshot workflow in\n  [`TASKS.md ## test`](TASKS.md#test) and consumed by\n  [`doca-debug ## test`](..\u002F..\u002Fdoca-debug\u002FTASKS.md#test) step 3\n  (read-only triple) and\n  [`doca-programming-guide ## debug`](..\u002F..\u002Fdoca-programming-guide\u002FTASKS.md#debug).\n- **\"`doca_caps` returned nothing for capability Y — what does that\n  mean?\"** — worked example: *\"empty output for RDMA\"*. Answered by\n  the empty-output interpretation rules in\n  [`TASKS.md ## debug`](TASKS.md#debug) +\n  [`CAPABILITIES.md ## Error taxonomy`](CAPABILITIES.md#error-taxonomy).\n\n## Audience\n\nThis skill serves **external operators, developers, and AI agents who\nneed a side-effect-free way to ask \"what does DOCA see on this host?\"**\nbefore doing anything that changes state. Concretely:\n\n- An external developer who installed DOCA (or is using the public NGC\n  DOCA container per [`doca-setup ## no-install`](..\u002F..\u002Fdoca-setup\u002FTASKS.md#no-install))\n  and wants to confirm the install can see hardware before writing code.\n- A platform operator deploying a DOCA service who wants a documented,\n  read-only snapshot of *which DOCA libraries this host actually\n  supports* and *which capabilities each DOCA device offers*.\n- An AI agent producing a *capability snapshot* artifact during the\n  documented setup or programming-guide debug procedures (it's listed\n  as the canonical first step in\n  [`doca-setup ## test`](..\u002F..\u002Fdoca-setup\u002FTASKS.md#test) and\n  [`doca-programming-guide ## debug`](..\u002F..\u002Fdoca-programming-guide\u002FTASKS.md#debug)).\n\nIt is **not** for users debugging `doca_caps` itself, and **not** a\nsubstitute for the live public Capabilities Print Tool guide.\n\n`doca_caps` is shipped as a **tool** (a single CLI binary), not a\nlibrary you link against. The skill uses the same `kind: tool`\nthree-file shape as the rest of the bundle so the agent's task-verb\ncontract (`configure \u002F build \u002F modify \u002F run \u002F test \u002F debug`) is uniform\nacross libraries, services, and tools — even when individual verbs\ncollapse to a routing stub for a shipped read-only binary.\n\n## When to load this skill\n\nLoad this skill when the user is — or the agent needs to — invoke\n`doca_caps` on a real host with DOCA installed (or inside the public\nNGC DOCA container). Concretely:\n\n- Running `doca_caps --list-devs` to enumerate DOCA devices.\n- Running `doca_caps --list-rep-devs` to enumerate representor\n  devices.\n- Scoping output to a specific PCIe address with `--pci-addr`.\n- Listing the DOCA libraries the install reports as available on the\n  current OS.\n- Listing the available DOCA logger names.\n- Capturing a documented, side-effect-free **capability snapshot** as\n  prerequisite evidence for later `## debug` workflows.\n\nDo **not** load this skill for general DOCA orientation, library API\nwork, or installation. For those, use\n[`doca-public-knowledge-map`](..\u002F..\u002Fdoca-public-knowledge-map\u002FSKILL.md),\nthe matching `libs\u002F\u003Clibrary>` skill, or\n[`doca-setup`](..\u002F..\u002Fdoca-setup\u002FSKILL.md).\n\n## What this skill provides\n\nThis is a **thin loader**. Substantive material lives in two companion\nfiles:\n\n- `CAPABILITIES.md` — what `doca_caps` reports (the five documented\n  capability families: devices, representors, libraries, library\n  capabilities, loggers), version availability and execution\n  environment, the tool's narrow error surface, its observability role\n  inside other skills' workflows, and its read-only safety posture.\n- `TASKS.md` — step-by-step workflows for the in-scope task verbs:\n  `configure` (route to install), `build` (route to install), `modify`\n  (refuse), `run` (the documented invocations), `test` (capability\n  snapshot as install smoke-test), `debug` (what to do when the tool\n  reports nothing or fails), plus a `Deferred task verbs` block routing\n  out-of-scope questions.\n\nThe skill assumes a host where DOCA is already installed (or the public\nNGC DOCA container is running) and the operator has whatever\npermissions the public guide requires for `doca_caps` to enumerate\ndevices on their platform.\n\n## What this skill deliberately does not ship\n\nThis skill is **agent guidance**, not a samples or scripts bundle. To\nkeep the boundary clean, it deliberately does not contain — and pull\nrequests should not add:\n\n- **Pre-baked example output.** Output is install- and\n  hardware-specific. A captured example pinned to one platform and\n  one DOCA version misleads operators on a different platform \u002F version.\n- **Wrappers, parsers, or scripts** in any language that consume\n  `doca_caps` output. The output format is documented; if a user\n  wants to script against it, the right answer is \"read the live\n  guide, write the parser against your installed version\".\n- **A `samples\u002F` or `reference\u002F` subtree.** This is a thin loader for\n  a documented CLI; substantive material lives on the public page and\n  in `--help`.\n\n## Loading order\n\n1. Read this `SKILL.md` first to confirm the user's question is in\n   scope (the user actually wants to invoke `doca_caps`, not learn\n   about DOCA in general).\n2. **For what `doca_caps` reports, version availability, error\n   surface, and safety posture, see [CAPABILITIES.md](CAPABILITIES.md).**\n3. **For the documented invocations and the capability-snapshot\n   workflow — `configure`, `build`, `modify`, `run`, `test`, `debug` —\n   see [TASKS.md](TASKS.md).**\n\n## Related skills\n\n- [`doca-public-knowledge-map`](..\u002F..\u002Fdoca-public-knowledge-map\u002FSKILL.md)\n  — routing to the public Capabilities Print Tool guide and the rest\n  of the public DOCA documentation set.\n- [`doca-setup`](..\u002F..\u002Fdoca-setup\u002FSKILL.md) — env preparation, install\n  verification (`doca_caps` is the canonical first step there), and\n  the *I have no install yet* path with the public NGC DOCA container.\n- [`doca-programming-guide`](..\u002F..\u002Fdoca-programming-guide\u002FSKILL.md) —\n  cross-library programming patterns, including the `## debug`\n  procedure where the saved `doca_caps` snapshot is consumed.\n- The matching `libs\u002F\u003Clibrary>` skill — for fine-grained,\n  library-specific capability questions that go beyond what\n  `doca_caps` exposes.\n",{"data":32,"body":36},{"license":23,"name":4,"description":6,"metadata":33,"compatibility":35},{"kind":34},"tool","Requires DOCA SDK ≥ 2.6.0 installed at \u002Fopt\u002Fmellanox\u002Fdoca on Linux (Ubuntu 22.04\u002F24.04 or RHEL\u002FSLES) with a BlueField DPU or ConnectX NIC; runs identically on the host or on the BlueField Arm side. Invokes \u002Fopt\u002Fmellanox\u002Fdoca\u002Ftools\u002Fdoca_caps and reads `pkg-config doca-common` to confirm the install.\n",{"type":37,"children":38},"root",[39,57,157,164,176,470,476,488,560,585,617,623,635,703,743,749,761,846,858,864,876,939,945,1040,1046],{"type":40,"tag":41,"props":42,"children":44},"element","h1",{"id":43},"doca-capabilities-print-tool-doca_caps",[45,48,55],{"type":46,"value":47},"text","DOCA Capabilities Print Tool (",{"type":40,"tag":49,"props":50,"children":52},"code",{"className":51},[],[53],{"type":46,"value":54},"doca_caps",{"type":46,"value":56},")",{"type":40,"tag":58,"props":59,"children":60},"p",{},[61,67,69,74,76,86,88,98,100,110,112,117,119,128,130,143,145,155],{"type":40,"tag":62,"props":63,"children":64},"strong",{},[65],{"type":46,"value":66},"Where to start:",{"type":46,"value":68}," This is a tool skill for invoking ",{"type":40,"tag":49,"props":70,"children":72},{"className":71},[],[73],{"type":46,"value":54},{"type":46,"value":75},",\na side-effect-free CLI. Open ",{"type":40,"tag":77,"props":78,"children":80},"a",{"href":79},"TASKS.md",[81],{"type":40,"tag":49,"props":82,"children":84},{"className":83},[],[85],{"type":46,"value":79},{"type":46,"value":87}," and start at\n",{"type":40,"tag":77,"props":89,"children":91},{"href":90},"TASKS.md#run",[92],{"type":40,"tag":49,"props":93,"children":95},{"className":94},[],[96],{"type":46,"value":97},"## run",{"type":46,"value":99}," for the documented invocations, or\n",{"type":40,"tag":77,"props":101,"children":103},{"href":102},"TASKS.md#test",[104],{"type":40,"tag":49,"props":105,"children":107},{"className":106},[],[108],{"type":46,"value":109},"## test",{"type":46,"value":111}," when using ",{"type":40,"tag":49,"props":113,"children":115},{"className":114},[],[116],{"type":46,"value":54},{"type":46,"value":118}," as an install\nsmoke-test. Open ",{"type":40,"tag":77,"props":120,"children":122},{"href":121},"CAPABILITIES.md",[123],{"type":40,"tag":49,"props":124,"children":126},{"className":125},[],[127],{"type":46,"value":121},{"type":46,"value":129}," when the\nquestion is ",{"type":40,"tag":131,"props":132,"children":133},"em",{},[134,136,141],{"type":46,"value":135},"what kinds of capability families ",{"type":40,"tag":49,"props":137,"children":139},{"className":138},[],[140],{"type":46,"value":54},{"type":46,"value":142}," reports",{"type":46,"value":144},".\nIf DOCA is not installed yet, route to\n",{"type":40,"tag":77,"props":146,"children":148},{"href":147},"..\u002F..\u002Fdoca-setup\u002FSKILL.md",[149],{"type":40,"tag":49,"props":150,"children":152},{"className":151},[],[153],{"type":46,"value":154},"doca-setup",{"type":46,"value":156}," first.",{"type":40,"tag":158,"props":159,"children":161},"h2",{"id":160},"example-questions-this-skill-answers-well",[162],{"type":46,"value":163},"Example questions this skill answers well",{"type":40,"tag":58,"props":165,"children":166},{},[167,169,174],{"type":46,"value":168},"The CLASSES of ",{"type":40,"tag":49,"props":170,"children":172},{"className":171},[],[173],{"type":46,"value":54},{"type":46,"value":175}," questions this skill is built to answer,\neach with one worked example. The class is the load-bearing piece;\nthe worked example is one instance.",{"type":40,"tag":177,"props":178,"children":179},"ul",{},[180,235,277,325,373,424],{"type":40,"tag":181,"props":182,"children":183},"li",{},[184,189,191,196,198,208],{"type":40,"tag":62,"props":185,"children":186},{},[187],{"type":46,"value":188},"\"What DOCA devices does this host see?\"",{"type":46,"value":190}," — worked example: ",{"type":40,"tag":131,"props":192,"children":193},{},[194],{"type":46,"value":195},"\"is\nmy BlueField PF visible to DOCA\"",{"type":46,"value":197},". Answered by the device\nenumeration in\n",{"type":40,"tag":77,"props":199,"children":201},{"href":200},"CAPABILITIES.md#capabilities-and-modes",[202],{"type":40,"tag":49,"props":203,"children":205},{"className":204},[],[206],{"type":46,"value":207},"CAPABILITIES.md ## Capabilities and modes",{"type":40,"tag":177,"props":209,"children":210},{},[211],{"type":40,"tag":181,"props":212,"children":213},{},[214,216,222,224,233],{"type":46,"value":215},"the ",{"type":40,"tag":49,"props":217,"children":219},{"className":218},[],[220],{"type":46,"value":221},"--list-devs",{"type":46,"value":223}," invocation in\n",{"type":40,"tag":77,"props":225,"children":226},{"href":90},[227],{"type":40,"tag":49,"props":228,"children":230},{"className":229},[],[231],{"type":46,"value":232},"TASKS.md ## run",{"type":46,"value":234},".",{"type":40,"tag":181,"props":236,"children":237},{},[238,243,245,250,252,260],{"type":40,"tag":62,"props":239,"children":240},{},[241],{"type":46,"value":242},"\"Which DOCA libraries are available on this OS?\"",{"type":46,"value":244}," — worked\nexample: ",{"type":40,"tag":131,"props":246,"children":247},{},[248],{"type":46,"value":249},"\"is Flow available on my RHEL host\"",{"type":46,"value":251},". Answered by the\nlibrary-availability surface in\n",{"type":40,"tag":77,"props":253,"children":254},{"href":200},[255],{"type":40,"tag":49,"props":256,"children":258},{"className":257},[],[259],{"type":46,"value":207},{"type":40,"tag":177,"props":261,"children":262},{},[263],{"type":40,"tag":181,"props":264,"children":265},{},[266,268,276],{"type":46,"value":267},"the library-listing invocation in\n",{"type":40,"tag":77,"props":269,"children":270},{"href":90},[271],{"type":40,"tag":49,"props":272,"children":274},{"className":273},[],[275],{"type":46,"value":232},{"type":46,"value":234},{"type":40,"tag":181,"props":278,"children":279},{},[280,285,286,291,293,301],{"type":40,"tag":62,"props":281,"children":282},{},[283],{"type":46,"value":284},"\"Does this device support library X capability Y?\"",{"type":46,"value":244},{"type":40,"tag":131,"props":287,"children":288},{},[289],{"type":46,"value":290},"\"does this device support Flow hairpin?\"",{"type":46,"value":292},". Answered by\nthe per-device per-library capability matrix in\n",{"type":40,"tag":77,"props":294,"children":295},{"href":200},[296],{"type":40,"tag":49,"props":297,"children":299},{"className":298},[],[300],{"type":46,"value":207},{"type":40,"tag":177,"props":302,"children":303},{},[304],{"type":40,"tag":181,"props":305,"children":306},{},[307,308,314,316,324],{"type":46,"value":215},{"type":40,"tag":49,"props":309,"children":311},{"className":310},[],[312],{"type":46,"value":313},"--pci-addr",{"type":46,"value":315},"-scoped invocation in\n",{"type":40,"tag":77,"props":317,"children":318},{"href":90},[319],{"type":40,"tag":49,"props":320,"children":322},{"className":321},[],[323],{"type":46,"value":232},{"type":46,"value":234},{"type":40,"tag":181,"props":326,"children":327},{},[328,333,335,340,342,350],{"type":40,"tag":62,"props":329,"children":330},{},[331],{"type":46,"value":332},"\"What representors are visible to DOCA?\"",{"type":46,"value":334}," — worked example:\n",{"type":40,"tag":131,"props":336,"children":337},{},[338],{"type":46,"value":339},"\"enumerate VF representors for pf0\"",{"type":46,"value":341},". Answered by the representor\nenumeration in\n",{"type":40,"tag":77,"props":343,"children":344},{"href":200},[345],{"type":40,"tag":49,"props":346,"children":348},{"className":347},[],[349],{"type":46,"value":207},{"type":40,"tag":177,"props":351,"children":352},{},[353],{"type":40,"tag":181,"props":354,"children":355},{},[356,357,363,364,372],{"type":46,"value":215},{"type":40,"tag":49,"props":358,"children":360},{"className":359},[],[361],{"type":46,"value":362},"--list-rep-devs",{"type":46,"value":223},{"type":40,"tag":77,"props":365,"children":366},{"href":90},[367],{"type":40,"tag":49,"props":368,"children":370},{"className":369},[],[371],{"type":46,"value":232},{"type":46,"value":234},{"type":40,"tag":181,"props":374,"children":375},{},[376,381,383,388,390,399,401,411,413,423],{"type":40,"tag":62,"props":377,"children":378},{},[379],{"type":46,"value":380},"\"I want a snapshot of state to attach to my debug session.\"",{"type":46,"value":382}," —\nworked example: ",{"type":40,"tag":131,"props":384,"children":385},{},[386],{"type":46,"value":387},"\"save device + library + capability output to a\nfile\"",{"type":46,"value":389},". Answered by the snapshot workflow in\n",{"type":40,"tag":77,"props":391,"children":392},{"href":102},[393],{"type":40,"tag":49,"props":394,"children":396},{"className":395},[],[397],{"type":46,"value":398},"TASKS.md ## test",{"type":46,"value":400}," and consumed by\n",{"type":40,"tag":77,"props":402,"children":404},{"href":403},"..\u002F..\u002Fdoca-debug\u002FTASKS.md#test",[405],{"type":40,"tag":49,"props":406,"children":408},{"className":407},[],[409],{"type":46,"value":410},"doca-debug ## test",{"type":46,"value":412}," step 3\n(read-only triple) and\n",{"type":40,"tag":77,"props":414,"children":416},{"href":415},"..\u002F..\u002Fdoca-programming-guide\u002FTASKS.md#debug",[417],{"type":40,"tag":49,"props":418,"children":420},{"className":419},[],[421],{"type":46,"value":422},"doca-programming-guide ## debug",{"type":46,"value":234},{"type":40,"tag":181,"props":425,"children":426},{},[427,439,440,445,447,457,459,469],{"type":40,"tag":62,"props":428,"children":429},{},[430,432,437],{"type":46,"value":431},"\"",{"type":40,"tag":49,"props":433,"children":435},{"className":434},[],[436],{"type":46,"value":54},{"type":46,"value":438}," returned nothing for capability Y — what does that\nmean?\"",{"type":46,"value":190},{"type":40,"tag":131,"props":441,"children":442},{},[443],{"type":46,"value":444},"\"empty output for RDMA\"",{"type":46,"value":446},". Answered by\nthe empty-output interpretation rules in\n",{"type":40,"tag":77,"props":448,"children":450},{"href":449},"TASKS.md#debug",[451],{"type":40,"tag":49,"props":452,"children":454},{"className":453},[],[455],{"type":46,"value":456},"TASKS.md ## debug",{"type":46,"value":458}," +\n",{"type":40,"tag":77,"props":460,"children":462},{"href":461},"CAPABILITIES.md#error-taxonomy",[463],{"type":40,"tag":49,"props":464,"children":466},{"className":465},[],[467],{"type":46,"value":468},"CAPABILITIES.md ## Error taxonomy",{"type":46,"value":234},{"type":40,"tag":158,"props":471,"children":473},{"id":472},"audience",[474],{"type":46,"value":475},"Audience",{"type":40,"tag":58,"props":477,"children":478},{},[479,481,486],{"type":46,"value":480},"This skill serves ",{"type":40,"tag":62,"props":482,"children":483},{},[484],{"type":46,"value":485},"external operators, developers, and AI agents who\nneed a side-effect-free way to ask \"what does DOCA see on this host?\"",{"type":46,"value":487},"\nbefore doing anything that changes state. Concretely:",{"type":40,"tag":177,"props":489,"children":490},{},[491,508,526],{"type":40,"tag":181,"props":492,"children":493},{},[494,496,506],{"type":46,"value":495},"An external developer who installed DOCA (or is using the public NGC\nDOCA container per ",{"type":40,"tag":77,"props":497,"children":499},{"href":498},"..\u002F..\u002Fdoca-setup\u002FTASKS.md#no-install",[500],{"type":40,"tag":49,"props":501,"children":503},{"className":502},[],[504],{"type":46,"value":505},"doca-setup ## no-install",{"type":46,"value":507},")\nand wants to confirm the install can see hardware before writing code.",{"type":40,"tag":181,"props":509,"children":510},{},[511,513,518,520,525],{"type":46,"value":512},"A platform operator deploying a DOCA service who wants a documented,\nread-only snapshot of ",{"type":40,"tag":131,"props":514,"children":515},{},[516],{"type":46,"value":517},"which DOCA libraries this host actually\nsupports",{"type":46,"value":519}," and ",{"type":40,"tag":131,"props":521,"children":522},{},[523],{"type":46,"value":524},"which capabilities each DOCA device offers",{"type":46,"value":234},{"type":40,"tag":181,"props":527,"children":528},{},[529,531,536,538,548,550,558],{"type":46,"value":530},"An AI agent producing a ",{"type":40,"tag":131,"props":532,"children":533},{},[534],{"type":46,"value":535},"capability snapshot",{"type":46,"value":537}," artifact during the\ndocumented setup or programming-guide debug procedures (it's listed\nas the canonical first step in\n",{"type":40,"tag":77,"props":539,"children":541},{"href":540},"..\u002F..\u002Fdoca-setup\u002FTASKS.md#test",[542],{"type":40,"tag":49,"props":543,"children":545},{"className":544},[],[546],{"type":46,"value":547},"doca-setup ## test",{"type":46,"value":549}," and\n",{"type":40,"tag":77,"props":551,"children":552},{"href":415},[553],{"type":40,"tag":49,"props":554,"children":556},{"className":555},[],[557],{"type":46,"value":422},{"type":46,"value":559},").",{"type":40,"tag":58,"props":561,"children":562},{},[563,565,570,572,577,579,583],{"type":46,"value":564},"It is ",{"type":40,"tag":62,"props":566,"children":567},{},[568],{"type":46,"value":569},"not",{"type":46,"value":571}," for users debugging ",{"type":40,"tag":49,"props":573,"children":575},{"className":574},[],[576],{"type":46,"value":54},{"type":46,"value":578}," itself, and ",{"type":40,"tag":62,"props":580,"children":581},{},[582],{"type":46,"value":569},{"type":46,"value":584}," a\nsubstitute for the live public Capabilities Print Tool guide.",{"type":40,"tag":58,"props":586,"children":587},{},[588,593,595,599,601,607,609,615],{"type":40,"tag":49,"props":589,"children":591},{"className":590},[],[592],{"type":46,"value":54},{"type":46,"value":594}," is shipped as a ",{"type":40,"tag":62,"props":596,"children":597},{},[598],{"type":46,"value":34},{"type":46,"value":600}," (a single CLI binary), not a\nlibrary you link against. The skill uses the same ",{"type":40,"tag":49,"props":602,"children":604},{"className":603},[],[605],{"type":46,"value":606},"kind: tool",{"type":46,"value":608},"\nthree-file shape as the rest of the bundle so the agent's task-verb\ncontract (",{"type":40,"tag":49,"props":610,"children":612},{"className":611},[],[613],{"type":46,"value":614},"configure \u002F build \u002F modify \u002F run \u002F test \u002F debug",{"type":46,"value":616},") is uniform\nacross libraries, services, and tools — even when individual verbs\ncollapse to a routing stub for a shipped read-only binary.",{"type":40,"tag":158,"props":618,"children":620},{"id":619},"when-to-load-this-skill",[621],{"type":46,"value":622},"When to load this skill",{"type":40,"tag":58,"props":624,"children":625},{},[626,628,633],{"type":46,"value":627},"Load this skill when the user is — or the agent needs to — invoke\n",{"type":40,"tag":49,"props":629,"children":631},{"className":630},[],[632],{"type":46,"value":54},{"type":46,"value":634}," on a real host with DOCA installed (or inside the public\nNGC DOCA container). Concretely:",{"type":40,"tag":177,"props":636,"children":637},{},[638,651,663,674,679,684],{"type":40,"tag":181,"props":639,"children":640},{},[641,643,649],{"type":46,"value":642},"Running ",{"type":40,"tag":49,"props":644,"children":646},{"className":645},[],[647],{"type":46,"value":648},"doca_caps --list-devs",{"type":46,"value":650}," to enumerate DOCA devices.",{"type":40,"tag":181,"props":652,"children":653},{},[654,655,661],{"type":46,"value":642},{"type":40,"tag":49,"props":656,"children":658},{"className":657},[],[659],{"type":46,"value":660},"doca_caps --list-rep-devs",{"type":46,"value":662}," to enumerate representor\ndevices.",{"type":40,"tag":181,"props":664,"children":665},{},[666,668,673],{"type":46,"value":667},"Scoping output to a specific PCIe address with ",{"type":40,"tag":49,"props":669,"children":671},{"className":670},[],[672],{"type":46,"value":313},{"type":46,"value":234},{"type":40,"tag":181,"props":675,"children":676},{},[677],{"type":46,"value":678},"Listing the DOCA libraries the install reports as available on the\ncurrent OS.",{"type":40,"tag":181,"props":680,"children":681},{},[682],{"type":46,"value":683},"Listing the available DOCA logger names.",{"type":40,"tag":181,"props":685,"children":686},{},[687,689,693,695,701],{"type":46,"value":688},"Capturing a documented, side-effect-free ",{"type":40,"tag":62,"props":690,"children":691},{},[692],{"type":46,"value":535},{"type":46,"value":694}," as\nprerequisite evidence for later ",{"type":40,"tag":49,"props":696,"children":698},{"className":697},[],[699],{"type":46,"value":700},"## debug",{"type":46,"value":702}," workflows.",{"type":40,"tag":58,"props":704,"children":705},{},[706,708,712,714,724,726,732,734,742],{"type":46,"value":707},"Do ",{"type":40,"tag":62,"props":709,"children":710},{},[711],{"type":46,"value":569},{"type":46,"value":713}," load this skill for general DOCA orientation, library API\nwork, or installation. For those, use\n",{"type":40,"tag":77,"props":715,"children":717},{"href":716},"..\u002F..\u002Fdoca-public-knowledge-map\u002FSKILL.md",[718],{"type":40,"tag":49,"props":719,"children":721},{"className":720},[],[722],{"type":46,"value":723},"doca-public-knowledge-map",{"type":46,"value":725},",\nthe matching ",{"type":40,"tag":49,"props":727,"children":729},{"className":728},[],[730],{"type":46,"value":731},"libs\u002F\u003Clibrary>",{"type":46,"value":733}," skill, or\n",{"type":40,"tag":77,"props":735,"children":736},{"href":147},[737],{"type":40,"tag":49,"props":738,"children":740},{"className":739},[],[741],{"type":46,"value":154},{"type":46,"value":234},{"type":40,"tag":158,"props":744,"children":746},{"id":745},"what-this-skill-provides",[747],{"type":46,"value":748},"What this skill provides",{"type":40,"tag":58,"props":750,"children":751},{},[752,754,759],{"type":46,"value":753},"This is a ",{"type":40,"tag":62,"props":755,"children":756},{},[757],{"type":46,"value":758},"thin loader",{"type":46,"value":760},". Substantive material lives in two companion\nfiles:",{"type":40,"tag":177,"props":762,"children":763},{},[764,781],{"type":40,"tag":181,"props":765,"children":766},{},[767,772,774,779],{"type":40,"tag":49,"props":768,"children":770},{"className":769},[],[771],{"type":46,"value":121},{"type":46,"value":773}," — what ",{"type":40,"tag":49,"props":775,"children":777},{"className":776},[],[778],{"type":46,"value":54},{"type":46,"value":780}," reports (the five documented\ncapability families: devices, representors, libraries, library\ncapabilities, loggers), version availability and execution\nenvironment, the tool's narrow error surface, its observability role\ninside other skills' workflows, and its read-only safety posture.",{"type":40,"tag":181,"props":782,"children":783},{},[784,789,791,797,799,805,806,812,814,820,822,828,830,836,838,844],{"type":40,"tag":49,"props":785,"children":787},{"className":786},[],[788],{"type":46,"value":79},{"type":46,"value":790}," — step-by-step workflows for the in-scope task verbs:\n",{"type":40,"tag":49,"props":792,"children":794},{"className":793},[],[795],{"type":46,"value":796},"configure",{"type":46,"value":798}," (route to install), ",{"type":40,"tag":49,"props":800,"children":802},{"className":801},[],[803],{"type":46,"value":804},"build",{"type":46,"value":798},{"type":40,"tag":49,"props":807,"children":809},{"className":808},[],[810],{"type":46,"value":811},"modify",{"type":46,"value":813},"\n(refuse), ",{"type":40,"tag":49,"props":815,"children":817},{"className":816},[],[818],{"type":46,"value":819},"run",{"type":46,"value":821}," (the documented invocations), ",{"type":40,"tag":49,"props":823,"children":825},{"className":824},[],[826],{"type":46,"value":827},"test",{"type":46,"value":829}," (capability\nsnapshot as install smoke-test), ",{"type":40,"tag":49,"props":831,"children":833},{"className":832},[],[834],{"type":46,"value":835},"debug",{"type":46,"value":837}," (what to do when the tool\nreports nothing or fails), plus a ",{"type":40,"tag":49,"props":839,"children":841},{"className":840},[],[842],{"type":46,"value":843},"Deferred task verbs",{"type":46,"value":845}," block routing\nout-of-scope questions.",{"type":40,"tag":58,"props":847,"children":848},{},[849,851,856],{"type":46,"value":850},"The skill assumes a host where DOCA is already installed (or the public\nNGC DOCA container is running) and the operator has whatever\npermissions the public guide requires for ",{"type":40,"tag":49,"props":852,"children":854},{"className":853},[],[855],{"type":46,"value":54},{"type":46,"value":857}," to enumerate\ndevices on their platform.",{"type":40,"tag":158,"props":859,"children":861},{"id":860},"what-this-skill-deliberately-does-not-ship",[862],{"type":46,"value":863},"What this skill deliberately does not ship",{"type":40,"tag":58,"props":865,"children":866},{},[867,869,874],{"type":46,"value":868},"This skill is ",{"type":40,"tag":62,"props":870,"children":871},{},[872],{"type":46,"value":873},"agent guidance",{"type":46,"value":875},", not a samples or scripts bundle. To\nkeep the boundary clean, it deliberately does not contain — and pull\nrequests should not add:",{"type":40,"tag":177,"props":877,"children":878},{},[879,889,906],{"type":40,"tag":181,"props":880,"children":881},{},[882,887],{"type":40,"tag":62,"props":883,"children":884},{},[885],{"type":46,"value":886},"Pre-baked example output.",{"type":46,"value":888}," Output is install- and\nhardware-specific. A captured example pinned to one platform and\none DOCA version misleads operators on a different platform \u002F version.",{"type":40,"tag":181,"props":890,"children":891},{},[892,897,899,904],{"type":40,"tag":62,"props":893,"children":894},{},[895],{"type":46,"value":896},"Wrappers, parsers, or scripts",{"type":46,"value":898}," in any language that consume\n",{"type":40,"tag":49,"props":900,"children":902},{"className":901},[],[903],{"type":46,"value":54},{"type":46,"value":905}," output. The output format is documented; if a user\nwants to script against it, the right answer is \"read the live\nguide, write the parser against your installed version\".",{"type":40,"tag":181,"props":907,"children":908},{},[909,930,932,938],{"type":40,"tag":62,"props":910,"children":911},{},[912,914,920,922,928],{"type":46,"value":913},"A ",{"type":40,"tag":49,"props":915,"children":917},{"className":916},[],[918],{"type":46,"value":919},"samples\u002F",{"type":46,"value":921}," or ",{"type":40,"tag":49,"props":923,"children":925},{"className":924},[],[926],{"type":46,"value":927},"reference\u002F",{"type":46,"value":929}," subtree.",{"type":46,"value":931}," This is a thin loader for\na documented CLI; substantive material lives on the public page and\nin ",{"type":40,"tag":49,"props":933,"children":935},{"className":934},[],[936],{"type":46,"value":937},"--help",{"type":46,"value":234},{"type":40,"tag":158,"props":940,"children":942},{"id":941},"loading-order",[943],{"type":46,"value":944},"Loading order",{"type":40,"tag":946,"props":947,"children":948},"ol",{},[949,969,989],{"type":40,"tag":181,"props":950,"children":951},{},[952,954,960,962,967],{"type":46,"value":953},"Read this ",{"type":40,"tag":49,"props":955,"children":957},{"className":956},[],[958],{"type":46,"value":959},"SKILL.md",{"type":46,"value":961}," first to confirm the user's question is in\nscope (the user actually wants to invoke ",{"type":40,"tag":49,"props":963,"children":965},{"className":964},[],[966],{"type":46,"value":54},{"type":46,"value":968},", not learn\nabout DOCA in general).",{"type":40,"tag":181,"props":970,"children":971},{},[972],{"type":40,"tag":62,"props":973,"children":974},{},[975,977,982,984,988],{"type":46,"value":976},"For what ",{"type":40,"tag":49,"props":978,"children":980},{"className":979},[],[981],{"type":46,"value":54},{"type":46,"value":983}," reports, version availability, error\nsurface, and safety posture, see ",{"type":40,"tag":77,"props":985,"children":986},{"href":121},[987],{"type":46,"value":121},{"type":46,"value":234},{"type":40,"tag":181,"props":990,"children":991},{},[992],{"type":40,"tag":62,"props":993,"children":994},{},[995,997,1002,1004,1009,1010,1015,1016,1021,1022,1027,1028,1033,1035,1039],{"type":46,"value":996},"For the documented invocations and the capability-snapshot\nworkflow — ",{"type":40,"tag":49,"props":998,"children":1000},{"className":999},[],[1001],{"type":46,"value":796},{"type":46,"value":1003},", ",{"type":40,"tag":49,"props":1005,"children":1007},{"className":1006},[],[1008],{"type":46,"value":804},{"type":46,"value":1003},{"type":40,"tag":49,"props":1011,"children":1013},{"className":1012},[],[1014],{"type":46,"value":811},{"type":46,"value":1003},{"type":40,"tag":49,"props":1017,"children":1019},{"className":1018},[],[1020],{"type":46,"value":819},{"type":46,"value":1003},{"type":40,"tag":49,"props":1023,"children":1025},{"className":1024},[],[1026],{"type":46,"value":827},{"type":46,"value":1003},{"type":40,"tag":49,"props":1029,"children":1031},{"className":1030},[],[1032],{"type":46,"value":835},{"type":46,"value":1034}," —\nsee ",{"type":40,"tag":77,"props":1036,"children":1037},{"href":79},[1038],{"type":46,"value":79},{"type":46,"value":234},{"type":40,"tag":158,"props":1041,"children":1043},{"id":1042},"related-skills",[1044],{"type":46,"value":1045},"Related skills",{"type":40,"tag":177,"props":1047,"children":1048},{},[1049,1062,1089,1118],{"type":40,"tag":181,"props":1050,"children":1051},{},[1052,1060],{"type":40,"tag":77,"props":1053,"children":1054},{"href":716},[1055],{"type":40,"tag":49,"props":1056,"children":1058},{"className":1057},[],[1059],{"type":46,"value":723},{"type":46,"value":1061},"\n— routing to the public Capabilities Print Tool guide and the rest\nof the public DOCA documentation set.",{"type":40,"tag":181,"props":1063,"children":1064},{},[1065,1073,1075,1080,1082,1087],{"type":40,"tag":77,"props":1066,"children":1067},{"href":147},[1068],{"type":40,"tag":49,"props":1069,"children":1071},{"className":1070},[],[1072],{"type":46,"value":154},{"type":46,"value":1074}," — env preparation, install\nverification (",{"type":40,"tag":49,"props":1076,"children":1078},{"className":1077},[],[1079],{"type":46,"value":54},{"type":46,"value":1081}," is the canonical first step there), and\nthe ",{"type":40,"tag":131,"props":1083,"children":1084},{},[1085],{"type":46,"value":1086},"I have no install yet",{"type":46,"value":1088}," path with the public NGC DOCA container.",{"type":40,"tag":181,"props":1090,"children":1091},{},[1092,1102,1104,1109,1111,1116],{"type":40,"tag":77,"props":1093,"children":1095},{"href":1094},"..\u002F..\u002Fdoca-programming-guide\u002FSKILL.md",[1096],{"type":40,"tag":49,"props":1097,"children":1099},{"className":1098},[],[1100],{"type":46,"value":1101},"doca-programming-guide",{"type":46,"value":1103}," —\ncross-library programming patterns, including the ",{"type":40,"tag":49,"props":1105,"children":1107},{"className":1106},[],[1108],{"type":46,"value":700},{"type":46,"value":1110},"\nprocedure where the saved ",{"type":40,"tag":49,"props":1112,"children":1114},{"className":1113},[],[1115],{"type":46,"value":54},{"type":46,"value":1117}," snapshot is consumed.",{"type":40,"tag":181,"props":1119,"children":1120},{},[1121,1123,1128,1130,1135],{"type":46,"value":1122},"The matching ",{"type":40,"tag":49,"props":1124,"children":1126},{"className":1125},[],[1127],{"type":46,"value":731},{"type":46,"value":1129}," skill — for fine-grained,\nlibrary-specific capability questions that go beyond what\n",{"type":40,"tag":49,"props":1131,"children":1133},{"className":1132},[],[1134],{"type":46,"value":54},{"type":46,"value":1136}," exposes.",{"items":1138,"total":1242},[1139,1156,1170,1184,1196,1213,1228],{"slug":1140,"name":1140,"fn":1141,"description":1142,"org":1143,"tags":1144,"stars":20,"repoUrl":21,"updatedAt":1155},"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},[1145,1148,1151,1152],{"name":1146,"slug":1147,"type":15},"Data Analysis","data-analysis",{"name":1149,"slug":1150,"type":15},"Data Engineering","data-engineering",{"name":9,"slug":8,"type":15},{"name":1153,"slug":1154,"type":15},"Performance","performance","2026-07-14T05:28:43.176466",{"slug":1157,"name":1157,"fn":1158,"description":1159,"org":1160,"tags":1161,"stars":20,"repoUrl":21,"updatedAt":1169},"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},[1162,1165,1168],{"name":1163,"slug":1164,"type":15},"Deployment","deployment",{"name":1166,"slug":1167,"type":15},"Infrastructure","infrastructure",{"name":9,"slug":8,"type":15},"2026-07-14T05:29:06.667109",{"slug":1171,"name":1171,"fn":1172,"description":1173,"org":1174,"tags":1175,"stars":20,"repoUrl":21,"updatedAt":1183},"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},[1176,1179,1180],{"name":1177,"slug":1178,"type":15},"Agents","agents",{"name":9,"slug":8,"type":15},{"name":1181,"slug":1182,"type":15},"Research","research","2026-07-14T05:28:06.816956",{"slug":1185,"name":1185,"fn":1186,"description":1187,"org":1188,"tags":1189,"stars":20,"repoUrl":21,"updatedAt":1195},"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},[1190,1191,1192],{"name":1146,"slug":1147,"type":15},{"name":9,"slug":8,"type":15},{"name":1193,"slug":1194,"type":15},"Testing","testing","2026-07-17T05:29:03.913266",{"slug":1197,"name":1197,"fn":1198,"description":1199,"org":1200,"tags":1201,"stars":20,"repoUrl":21,"updatedAt":1212},"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},[1202,1205,1208,1209],{"name":1203,"slug":1204,"type":15},"Automation","automation",{"name":1206,"slug":1207,"type":15},"Imaging","imaging",{"name":9,"slug":8,"type":15},{"name":1210,"slug":1211,"type":15},"Video","video","2026-07-17T05:28:53.905004",{"slug":1214,"name":1214,"fn":1215,"description":1216,"org":1217,"tags":1218,"stars":20,"repoUrl":21,"updatedAt":1227},"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},[1219,1220,1223,1224],{"name":1163,"slug":1164,"type":15},{"name":1221,"slug":1222,"type":15},"Docker","docker",{"name":9,"slug":8,"type":15},{"name":1225,"slug":1226,"type":15},"Operations","operations","2026-07-17T05:28:56.913999",{"slug":1229,"name":1229,"fn":1230,"description":1231,"org":1232,"tags":1233,"stars":20,"repoUrl":21,"updatedAt":1241},"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},[1234,1235,1238],{"name":9,"slug":8,"type":15},{"name":1236,"slug":1237,"type":15},"Quantum Computing","quantum-computing",{"name":1239,"slug":1240,"type":15},"Simulation","simulation","2026-07-14T05:26:58.898253",305,{"items":1244,"total":1395},[1245,1263,1279,1290,1302,1316,1329,1343,1354,1363,1377,1386],{"slug":1246,"name":1246,"fn":1247,"description":1248,"org":1249,"tags":1250,"stars":1260,"repoUrl":1261,"updatedAt":1262},"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},[1251,1254,1257],{"name":1252,"slug":1253,"type":15},"Documentation","documentation",{"name":1255,"slug":1256,"type":15},"MCP","mcp",{"name":1258,"slug":1259,"type":15},"Search","search",21777,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FNemoClaw","2026-07-20T06:00:01.461044",{"slug":1264,"name":1264,"fn":1265,"description":1266,"org":1267,"tags":1268,"stars":1276,"repoUrl":1277,"updatedAt":1278},"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},[1269,1272,1273],{"name":1270,"slug":1271,"type":15},"Containers","containers",{"name":1163,"slug":1164,"type":15},{"name":1274,"slug":1275,"type":15},"Python","python",17049,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FMegatron-LM","2026-07-27T06:06:11.249662",{"slug":1280,"name":1280,"fn":1281,"description":1282,"org":1283,"tags":1284,"stars":1276,"repoUrl":1277,"updatedAt":1289},"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},[1285,1288],{"name":1286,"slug":1287,"type":15},"CI\u002FCD","ci-cd",{"name":1163,"slug":1164,"type":15},"2026-07-14T05:25:59.97109",{"slug":1291,"name":1291,"fn":1292,"description":1293,"org":1294,"tags":1295,"stars":1276,"repoUrl":1277,"updatedAt":1301},"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},[1296,1297,1298],{"name":1286,"slug":1287,"type":15},{"name":1163,"slug":1164,"type":15},{"name":1299,"slug":1300,"type":15},"GitHub","github","2026-07-27T06:06:12.278222",{"slug":1303,"name":1303,"fn":1304,"description":1305,"org":1306,"tags":1307,"stars":1276,"repoUrl":1277,"updatedAt":1315},"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},[1308,1311,1312],{"name":1309,"slug":1310,"type":15},"Debugging","debugging",{"name":1299,"slug":1300,"type":15},{"name":1313,"slug":1314,"type":15},"Triage","triage","2026-07-14T05:25:57.442089",{"slug":1317,"name":1317,"fn":1318,"description":1319,"org":1320,"tags":1321,"stars":1276,"repoUrl":1277,"updatedAt":1328},"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},[1322,1325],{"name":1323,"slug":1324,"type":15},"Best Practices","best-practices",{"name":1326,"slug":1327,"type":15},"Code Analysis","code-analysis","2026-07-14T05:25:56.18433",{"slug":1330,"name":1330,"fn":1331,"description":1332,"org":1333,"tags":1334,"stars":1276,"repoUrl":1277,"updatedAt":1342},"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},[1335,1338,1341],{"name":1336,"slug":1337,"type":15},"Machine Learning","machine-learning",{"name":1339,"slug":1340,"type":15},"Migration","migration",{"name":9,"slug":8,"type":15},"2026-07-17T06:07:11.777011",{"slug":1344,"name":1344,"fn":1345,"description":1346,"org":1347,"tags":1348,"stars":1276,"repoUrl":1277,"updatedAt":1353},"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},[1349,1352],{"name":1350,"slug":1351,"type":15},"QA","qa",{"name":1193,"slug":1194,"type":15},"2026-07-14T05:25:53.673039",{"slug":1355,"name":1355,"fn":1356,"description":1357,"org":1358,"tags":1359,"stars":1276,"repoUrl":1277,"updatedAt":1362},"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},[1360,1361],{"name":1163,"slug":1164,"type":15},{"name":1166,"slug":1167,"type":15},"2026-07-14T05:25:49.362534",{"slug":1364,"name":1364,"fn":1365,"description":1366,"org":1367,"tags":1368,"stars":1276,"repoUrl":1277,"updatedAt":1376},"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},[1369,1372,1373],{"name":1370,"slug":1371,"type":15},"Code Review","code-review",{"name":1299,"slug":1300,"type":15},{"name":1374,"slug":1375,"type":15},"Pull Requests","pull-requests","2026-07-14T05:26:01.226578",{"slug":1378,"name":1378,"fn":1379,"description":1380,"org":1381,"tags":1382,"stars":1276,"repoUrl":1277,"updatedAt":1385},"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},[1383,1384],{"name":1350,"slug":1351,"type":15},{"name":1193,"slug":1194,"type":15},"2026-07-14T05:25:54.928983",{"slug":1387,"name":1387,"fn":1388,"description":1389,"org":1390,"tags":1391,"stars":1276,"repoUrl":1277,"updatedAt":1394},"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},[1392,1393],{"name":1203,"slug":1204,"type":15},{"name":1286,"slug":1287,"type":15},"2026-07-30T05:29:03.275638",496]