[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill-nvidia-doca-public-knowledge-map":3,"mdc-1a1v79-key":31,"related-org-nvidia-doca-public-knowledge-map":1198,"related-repo-nvidia-doca-public-knowledge-map":1357},{"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-public-knowledge-map","locate NVIDIA DOCA developer documentation","Use this skill when the user needs to locate authoritative information about NVIDIA DOCA without access to the source tree — finding the right docs.nvidia.com page for a library\u002Fservice\u002Ftool, identifying which DOCA libraries are installed and at what version, locating a sample on disk or its public GitHub source, decoding an on-disk path under \u002Fopt\u002Fmellanox\u002Fdoca, or recovering from a 404'd or renamed doc URL. Trigger even when the user does not explicitly mention 'DOCA' or 'docs.nvidia.com' — typical implicit phrasings include 'where can I read about this library', 'which version do I have installed', 'where is the sample for X', 'this NVIDIA URL is broken what is the new one', 'what is in \u002Fopt\u002Fmellanox\u002Fdoca', or 'where can I ask NVIDIA about this'. Refuse and route elsewhere for hands-on programming patterns, env prep and install verification, library API tutorials, or hardware\u002Ffirmware mutation — those belong to doca-programming-guide, doca-setup, the per-library skills, and doca-hardware-safety.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},"nvidia","NVIDIA","https:\u002F\u002Fpexgzepcugksgbtrxkhf.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Forg-logos\u002Fnvidia.png",[12,16,19],{"name":13,"slug":14,"type":15},"Documentation","documentation","tag",{"name":17,"slug":18,"type":15},"Reference","reference",{"name":9,"slug":8,"type":15},2473,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Fskills","2026-07-30T05:28:21.420404","Apache-2.0 AND CC-BY-4.0",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-public-knowledge-map","---\nlicense: Apache-2.0 AND CC-BY-4.0\nname: doca-public-knowledge-map\ndescription: >\n  Use this skill when the user needs to locate authoritative\n  information about NVIDIA DOCA without access to the source tree —\n  finding the right docs.nvidia.com page for a library\u002Fservice\u002Ftool,\n  identifying which DOCA libraries are installed and at what version,\n  locating a sample on disk or its public GitHub source, decoding an\n  on-disk path under \u002Fopt\u002Fmellanox\u002Fdoca, or recovering from a 404'd\n  or renamed doc URL. Trigger even when the user does not explicitly\n  mention 'DOCA' or 'docs.nvidia.com' — typical implicit phrasings\n  include 'where can I read about this library', 'which version do I\n  have installed', 'where is the sample for X', 'this NVIDIA URL is\n  broken what is the new one', 'what is in \u002Fopt\u002Fmellanox\u002Fdoca', or\n  'where can I ask NVIDIA about this'. Refuse and route elsewhere\n  for hands-on programming patterns, env prep and install\n  verification, library API tutorials, or hardware\u002Ffirmware\n  mutation — those belong to doca-programming-guide, doca-setup, the\n  per-library skills, and doca-hardware-safety.\nmetadata:\n  kind: knowledge\ncompatibility: >\n  No DOCA install required to read this skill (it is an overlay\n  loaded against any DOCA artifact skill); the validation steps\n  within this skill require a live DOCA install at \u002Fopt\u002Fmellanox\u002Fdoca.\n---\n\n# DOCA Public Knowledge Map\n\n**Where to start:** Reach for this skill whenever the question is \"where does\nthe authoritative answer live?\" — a docs page, the on-disk install layout, a\nsample, or an NGC catalog entry. Read [`## Public documentation entry points`](#public-documentation-entry-points)\nfirst; then jump to the routing-table section that matches the user's intent.\n\n## When to load this skill\n\nLoad this skill whenever the user asks anything about NVIDIA DOCA where the\nagent needs to **locate authoritative information** without access to the\nDOCA source tree. That includes **documentation-routing questions** about\ninstalling DOCA, building a sample, learning a DOCA library (Flow, DPA, Comm\nChannel, GPUNetIO, …), debugging an error, finding an API or error reference,\nfinding a sample, release notes, or the developer forum. This skill locates\nthe authoritative answer; hands-on installation, building, tutorials, and\ndebugging route to the workflow-owning skills named in the frontmatter and\n[`## Related skills`](#related-skills).\n\nThis skill is intentionally a **routing table**, not a tutorial. Pick the\nentry that matches the user's intent, fetch the URL or inspect the local\ninstall path, and only then answer.\n\n## First-contact discovery — the four questions to ask before any drill-down\n\nWhen a user opens with an open-ended orientation question (*\"I'm new with\nDOCA, how do I start?\"*, *\"can you guide me?\"*, *\"what's the easiest way to\ntry DOCA?\"*), the agent does **not** have enough information yet to pick a\npath. Asking these four questions before drilling avoids wasted recommendations\nthat the user cannot actually execute on their setup. Ask them as a single\nshort message; do not interrogate one-at-a-time.\n\n| Question | Why it matters | What it routes |\n| --- | --- | --- |\n| 1. **What OS are you on?** macOS, Windows, Linux laptop, cloud VM, lab Linux box, BlueField OS itself? | DOCA installs natively only on supported Linux distributions; macOS \u002F Windows users cannot install it at all. | Picks between the four DOCA acquisition paths in [`doca-setup ## no-install`](..\u002Fdoca-setup\u002FTASKS.md#no-install). macOS \u002F Windows \u002F no-Linux → Path 0 (NGC container). Supported Linux → Path A or B per the Installation Guide. |\n| 2. **What hardware do you have?** No NVIDIA hardware, ConnectX SmartNIC, BlueField as a SmartNIC in a host, BlueField as a standalone DPU, not sure? | Real-traffic runtime needs a real NIC; build \u002F read \u002F learn does not. The user's hardware decides which DOCA libraries are even relevant. | Picks the runtime story (container is build-only without hardware). Filters which libraries make sense to learn (Flow needs a real port to do anything visible; Comch needs a host ↔ DPU pair). |\n| 3. **What's your goal?** Just exploring, building a small first app on a specific library (Flow \u002F RDMA \u002F Comch \u002F Telemetry \u002F GPUNetIO \u002F DPA \u002F …), running an existing reference application, operating a service (DMS \u002F DTS \u002F BlueMan \u002F Firefly), or something else? | The bundle's first-app workflow (`doca-programming-guide ## modify`) starts from a **shipped C sample** and edits down. The right sample depends on the library the user is targeting. | Picks which library skill (if any) to load next. If the user does not yet know which library — that itself is a routing answer (see the *Library- and module-specific guides* table above and let the user pick). |\n| 4. **Which language do you plan to write the program in?** C \u002F C++, Rust, Go, Python, other? | DOCA's public surface is a C ABI. Non-C consumers go through FFI \u002F language bindings (`doca-programming-guide CAPABILITIES.md ## Capabilities and modes` and the per-library skill). The C samples are the reference even when the user's language is not C. | Picks whether the agent's first-app guidance is *direct C build* or *FFI \u002F bindings against the C ABI*. Does **not** change which sample the agent points at first. |\n\nThe four-question gate applies when an open-ended orientation, install,\nbuild, or run recommendation depends on the user's environment. For that\nclass, **never recommend a specific install path, container tag, or runnable\nsample without first having the answers to questions 1–3** (question 4 is\nneeded for the first-app workflow but not for orientation itself). Pure\ndocumentation lookups — a URL, API reference, sample location, version\nreference, or stale-link recovery — skip this gate and route directly.\nIf the user cannot or will not answer a required question, state which\nenvironment-dependent recommendation cannot be made, give only the safe\nversion-agnostic umbrella documentation route, and do not guess the missing\nvalue. Volunteering specific commands before this is the single most common\nfailure mode for DOCA orientation.\n\nIf the user has already volunteered some of the information in their first\nmessage, mark those questions answered and only ask the rest. Do not re-ask\nwhat the user has already told you.\n\n## Topic to \"where to look first\" routing table\n\nWhen the user asks something, route as follows:\n\n| User intent | First place to look |\n| --- | --- |\n| \"How do I install DOCA?\" | Installation Guide + Downloads page (Public documentation entry points). |\n| \"How do I start with DOCA — what's the very first thing?\" | Developer Quick Start Guide *if* the user has BlueField + host hardware; otherwise [`doca-setup ## no-install`](..\u002Fdoca-setup\u002FTASKS.md#no-install) Path 0 (NGC container). Use the four questions in [*First-contact discovery*](#first-contact-discovery--the-four-questions-to-ask-before-any-drill-down) to pick. |\n| \"Do I need a BlueField? A SmartNIC? Or just DOCA-Host?\" | Overview page (`doca-overview\u002Findex.html`) plus the Installation Guide's *DOCA installation profiles* section. The bundle does not pick the hardware for the user — these two pages do. |\n| \"Which package gives me library X?\" | Installation Guide section on package matrix; then verify on the user's system with `pkg-config --list-all`. |\n| \"Show me a sample that uses library X.\" | `\u002Fopt\u002Fmellanox\u002Fdoca\u002Fsamples\u002Fdoca_\u003CX>\u002F` if installed; otherwise the per-library guide on `docs.nvidia.com\u002Fdoca\u002Fsdk\u002F` (each library guide documents the samples shipped with it). |\n| \"How do I build a DOCA sample?\" | Library guide + the sample's own `meson.build` inside `\u002Fopt\u002Fmellanox\u002Fdoca\u002Fsamples\u002F...`. |\n| \"What is the API for X?\" | Library guide; confirm by inspecting headers under `$(pkg-config --variable=includedir doca-common)` (commonly `\u002Fopt\u002Fmellanox\u002Fdoca\u002Finclude\u002F` on DOCA 3.3+; `\u002Fopt\u002Fmellanox\u002Fdoca\u002Finfrastructure\u002Finclude\u002F` on legacy \u002F split-profile installs). |\n| \"Why does my build fail with `pkg-config` not finding `doca-...`?\" | \"Layout of an installed DOCA package\" (`PKG_CONFIG_PATH`), then load [`doca-setup`](..\u002Fdoca-setup\u002FSKILL.md) for environment preparation and debugging; do not silently mutate the environment. |\n| \"What is the latest version \u002F what changed?\" | Release Notes. |\n| \"Is there a newer DOCA — should I upgrade \u002F downgrade \u002F roll back?\" | [`doca-upgrade`](..\u002Fdoca-upgrade\u002FSKILL.md) for the detect → report → ASK → guided-upgrade discipline (never auto-upgrade); Release Notes for what the target release contains. |\n| \"Is the component I depend on being sunset \u002F deprecated?\" | Release Notes (lifecycle \u002F deprecation notices), reached via the [`doca-upgrade`](..\u002Fdoca-upgrade\u002FSKILL.md) sunset-awareness check; do not assert deprecation from memory. |\n| \"What does the DOCA version number mean? Is LTS still supported?\" | Compatibility Policy (Public documentation entry points table). |\n| \"How do I run DOCA on Kubernetes \u002F provision a DPU?\" | DOCA Platform Framework on GitHub. |\n| \"I have a behavior I cannot explain.\" | Release Notes (known issues) first; then the DOCA Developer Forum. Never go to the forum first. |\n| Any intent not matched above | Start at the DOCA SDK index and the Libraries \u002F Services \u002F Tools umbrella page in [`references\u002Fmap.md`](references\u002Fmap.md#public-documentation-entry-points). If the authoritative route remains unresolved, say so and use the Developer Forum; do not invent a URL or component name. |\n\n## What this skill deliberately does not cover\n\nThis file is intentionally a **map**, not a tutorial. It does not contain:\n\n- DOCA library tutorials (those live in the per-library guides).\n- API reference (lives in headers and the per-library guides).\n- Build-system deep-dives (lives in the Installation Guide and the sample\n  `meson.build` files).\n- Performance tuning, driver-level setup, OFED interaction (lives in the\n  Installation Guide and library-specific guides).\n\nWhen the agent needs those, it should fetch the matching public document or\nread the matching installed file. As more focused skills are added, they\nshould appear in [SKILLS.md](..\u002F..\u002FSKILLS.md) and link back here for the\n\"where to look\" lookups.\n\n## Related skills\n\nFor env preparation — install verification, build environment\n(`pkg-config`, headers, hugepages, devlink), env-class debugging, and\nthe *I have no install yet* path with the public NGC DOCA container\n(`nvcr.io\u002Fnvidia\u002Fdoca\u002Fdoca`) as the universal Stage-1 fallback for any\nuser on macOS, Windows, or Linux without DOCA — load\n[`doca-setup`](..\u002Fdoca-setup\u002FSKILL.md). That skill stops at *\"the\ninstall is healthy and the env is ready\"*.\n\nFor general DOCA programming patterns shared across every library —\nthe canonical `pkg-config doca-\u003Clibrary>` build pattern (C\u002FC++ direct\nor non-C via FFI), the universal *derive a custom first app from a\nshipped sample* workflow, the universal lifecycle (`cfg-create →\ninit → start → use → stop → destroy`), the cross-library\n`DOCA_ERROR_*` taxonomy, and the program-side debug order — load\n[`doca-programming-guide`](..\u002Fdoca-programming-guide\u002FSKILL.md). Each\nlibrary skill extends its `## modify` (first-app derivation) with\nlibrary-specific overrides.\n\nFor DOCA Flow internals — port and representor setup, pipe creation,\nmatch\u002Faction specifications, pipe validation before hardware programming,\nFlow counters and traces, Flow version compatibility, and debugging\n`DOCA_ERROR_*` failures from the Flow API — load\n[`doca-flow`](..\u002Flibs\u002Fdoca-flow\u002FSKILL.md). That skill assumes this one is\navailable for shared documentation routing and install-layout lookups,\n`doca-setup` for environment preparation, and `doca-programming-guide`\nfor the cross-library programming patterns it layers on top of.\n\n## Example questions this skill answers well\n\nSee [`references\u002Fmap.md`](references\u002Fmap.md#example-questions-this-skill-answers-well).\n## Library- and module-specific guides\n\nSee [`references\u002Fmap.md`](references\u002Fmap.md#library--and-module-specific-guides).\n## DOCA services\n\nSee [`references\u002Fmap.md`](references\u002Fmap.md#doca-services).\n## DOCA tools\n\nSee [`references\u002Fmap.md`](references\u002Fmap.md#doca-tools).\n## Externally-productized DOCA software — not in this bundle, but here is where to route\n\nSee [`references\u002Fmap.md`](references\u002Fmap.md#externally-productized-doca-software--not-in-this-bundle-but-here-is-where-to-route).\n## Public source code: GitHub\n\nSee [`references\u002Fmap.md`](references\u002Fmap.md#public-source-code-github).\n## Layout of an installed DOCA package\n\nSee [`references\u002Fmap.md`](references\u002Fmap.md#layout-of-an-installed-doca-package).\n## Where to find the version\n\nSee [`references\u002Fmap.md`](references\u002Fmap.md#where-to-find-the-version).\n## URL audit\n\nSee [`references\u002Fmap.md`](references\u002Fmap.md#url-audit).\n\n## Ground rules for any agent using this skill\n\nSee [`references\u002Fmap.md`](references\u002Fmap.md#ground-rules-for-any-agent-using-this-skill).\n## Public documentation entry points\n\nSee [`references\u002Fmap.md`](references\u002Fmap.md#public-documentation-entry-points).\n",{"data":32,"body":36},{"license":23,"name":4,"description":6,"metadata":33,"compatibility":35},{"kind":34},"knowledge","No DOCA install required to read this skill (it is an overlay loaded against any DOCA artifact skill); the validation steps within this skill require a live DOCA install at \u002Fopt\u002Fmellanox\u002Fdoca.\n",{"type":37,"children":38},"root",[39,47,73,80,111,123,129,162,357,369,374,380,385,773,779,791,823,836,842,885,941,979,985,1000,1005,1019,1025,1039,1045,1059,1065,1079,1085,1099,1105,1119,1125,1139,1145,1159,1165,1179,1185],{"type":40,"tag":41,"props":42,"children":43},"element","h1",{"id":4},[44],{"type":45,"value":46},"text","DOCA Public Knowledge Map",{"type":40,"tag":48,"props":49,"children":50},"p",{},[51,57,59,71],{"type":40,"tag":52,"props":53,"children":54},"strong",{},[55],{"type":45,"value":56},"Where to start:",{"type":45,"value":58}," Reach for this skill whenever the question is \"where does\nthe authoritative answer live?\" — a docs page, the on-disk install layout, a\nsample, or an NGC catalog entry. Read ",{"type":40,"tag":60,"props":61,"children":63},"a",{"href":62},"#public-documentation-entry-points",[64],{"type":40,"tag":65,"props":66,"children":68},"code",{"className":67},[],[69],{"type":45,"value":70},"## Public documentation entry points",{"type":45,"value":72},"\nfirst; then jump to the routing-table section that matches the user's intent.",{"type":40,"tag":74,"props":75,"children":77},"h2",{"id":76},"when-to-load-this-skill",[78],{"type":45,"value":79},"When to load this skill",{"type":40,"tag":48,"props":81,"children":82},{},[83,85,90,92,97,99,109],{"type":45,"value":84},"Load this skill whenever the user asks anything about NVIDIA DOCA where the\nagent needs to ",{"type":40,"tag":52,"props":86,"children":87},{},[88],{"type":45,"value":89},"locate authoritative information",{"type":45,"value":91}," without access to the\nDOCA source tree. That includes ",{"type":40,"tag":52,"props":93,"children":94},{},[95],{"type":45,"value":96},"documentation-routing questions",{"type":45,"value":98}," about\ninstalling DOCA, building a sample, learning a DOCA library (Flow, DPA, Comm\nChannel, GPUNetIO, …), debugging an error, finding an API or error reference,\nfinding a sample, release notes, or the developer forum. This skill locates\nthe authoritative answer; hands-on installation, building, tutorials, and\ndebugging route to the workflow-owning skills named in the frontmatter and\n",{"type":40,"tag":60,"props":100,"children":102},{"href":101},"#related-skills",[103],{"type":40,"tag":65,"props":104,"children":106},{"className":105},[],[107],{"type":45,"value":108},"## Related skills",{"type":45,"value":110},".",{"type":40,"tag":48,"props":112,"children":113},{},[114,116,121],{"type":45,"value":115},"This skill is intentionally a ",{"type":40,"tag":52,"props":117,"children":118},{},[119],{"type":45,"value":120},"routing table",{"type":45,"value":122},", not a tutorial. Pick the\nentry that matches the user's intent, fetch the URL or inspect the local\ninstall path, and only then answer.",{"type":40,"tag":74,"props":124,"children":126},{"id":125},"first-contact-discovery-the-four-questions-to-ask-before-any-drill-down",[127],{"type":45,"value":128},"First-contact discovery — the four questions to ask before any drill-down",{"type":40,"tag":48,"props":130,"children":131},{},[132,134,140,142,147,148,153,155,160],{"type":45,"value":133},"When a user opens with an open-ended orientation question (",{"type":40,"tag":135,"props":136,"children":137},"em",{},[138],{"type":45,"value":139},"\"I'm new with\nDOCA, how do I start?\"",{"type":45,"value":141},", ",{"type":40,"tag":135,"props":143,"children":144},{},[145],{"type":45,"value":146},"\"can you guide me?\"",{"type":45,"value":141},{"type":40,"tag":135,"props":149,"children":150},{},[151],{"type":45,"value":152},"\"what's the easiest way to\ntry DOCA?\"",{"type":45,"value":154},"), the agent does ",{"type":40,"tag":52,"props":156,"children":157},{},[158],{"type":45,"value":159},"not",{"type":45,"value":161}," have enough information yet to pick a\npath. Asking these four questions before drilling avoids wasted recommendations\nthat the user cannot actually execute on their setup. Ask them as a single\nshort message; do not interrogate one-at-a-time.",{"type":40,"tag":163,"props":164,"children":165},"table",{},[166,190],{"type":40,"tag":167,"props":168,"children":169},"thead",{},[170],{"type":40,"tag":171,"props":172,"children":173},"tr",{},[174,180,185],{"type":40,"tag":175,"props":176,"children":177},"th",{},[178],{"type":45,"value":179},"Question",{"type":40,"tag":175,"props":181,"children":182},{},[183],{"type":45,"value":184},"Why it matters",{"type":40,"tag":175,"props":186,"children":187},{},[188],{"type":45,"value":189},"What it routes",{"type":40,"tag":191,"props":192,"children":193},"tbody",{},[194,232,257,304],{"type":40,"tag":171,"props":195,"children":196},{},[197,210,215],{"type":40,"tag":198,"props":199,"children":200},"td",{},[201,203,208],{"type":45,"value":202},"1. ",{"type":40,"tag":52,"props":204,"children":205},{},[206],{"type":45,"value":207},"What OS are you on?",{"type":45,"value":209}," macOS, Windows, Linux laptop, cloud VM, lab Linux box, BlueField OS itself?",{"type":40,"tag":198,"props":211,"children":212},{},[213],{"type":45,"value":214},"DOCA installs natively only on supported Linux distributions; macOS \u002F Windows users cannot install it at all.",{"type":40,"tag":198,"props":216,"children":217},{},[218,220,230],{"type":45,"value":219},"Picks between the four DOCA acquisition paths in ",{"type":40,"tag":60,"props":221,"children":223},{"href":222},"..\u002Fdoca-setup\u002FTASKS.md#no-install",[224],{"type":40,"tag":65,"props":225,"children":227},{"className":226},[],[228],{"type":45,"value":229},"doca-setup ## no-install",{"type":45,"value":231},". macOS \u002F Windows \u002F no-Linux → Path 0 (NGC container). Supported Linux → Path A or B per the Installation Guide.",{"type":40,"tag":171,"props":233,"children":234},{},[235,247,252],{"type":40,"tag":198,"props":236,"children":237},{},[238,240,245],{"type":45,"value":239},"2. ",{"type":40,"tag":52,"props":241,"children":242},{},[243],{"type":45,"value":244},"What hardware do you have?",{"type":45,"value":246}," No NVIDIA hardware, ConnectX SmartNIC, BlueField as a SmartNIC in a host, BlueField as a standalone DPU, not sure?",{"type":40,"tag":198,"props":248,"children":249},{},[250],{"type":45,"value":251},"Real-traffic runtime needs a real NIC; build \u002F read \u002F learn does not. The user's hardware decides which DOCA libraries are even relevant.",{"type":40,"tag":198,"props":253,"children":254},{},[255],{"type":45,"value":256},"Picks the runtime story (container is build-only without hardware). Filters which libraries make sense to learn (Flow needs a real port to do anything visible; Comch needs a host ↔ DPU pair).",{"type":40,"tag":171,"props":258,"children":259},{},[260,272,292],{"type":40,"tag":198,"props":261,"children":262},{},[263,265,270],{"type":45,"value":264},"3. ",{"type":40,"tag":52,"props":266,"children":267},{},[268],{"type":45,"value":269},"What's your goal?",{"type":45,"value":271}," Just exploring, building a small first app on a specific library (Flow \u002F RDMA \u002F Comch \u002F Telemetry \u002F GPUNetIO \u002F DPA \u002F …), running an existing reference application, operating a service (DMS \u002F DTS \u002F BlueMan \u002F Firefly), or something else?",{"type":40,"tag":198,"props":273,"children":274},{},[275,277,283,285,290],{"type":45,"value":276},"The bundle's first-app workflow (",{"type":40,"tag":65,"props":278,"children":280},{"className":279},[],[281],{"type":45,"value":282},"doca-programming-guide ## modify",{"type":45,"value":284},") starts from a ",{"type":40,"tag":52,"props":286,"children":287},{},[288],{"type":45,"value":289},"shipped C sample",{"type":45,"value":291}," and edits down. The right sample depends on the library the user is targeting.",{"type":40,"tag":198,"props":293,"children":294},{},[295,297,302],{"type":45,"value":296},"Picks which library skill (if any) to load next. If the user does not yet know which library — that itself is a routing answer (see the ",{"type":40,"tag":135,"props":298,"children":299},{},[300],{"type":45,"value":301},"Library- and module-specific guides",{"type":45,"value":303}," table above and let the user pick).",{"type":40,"tag":171,"props":305,"children":306},{},[307,319,332],{"type":40,"tag":198,"props":308,"children":309},{},[310,312,317],{"type":45,"value":311},"4. ",{"type":40,"tag":52,"props":313,"children":314},{},[315],{"type":45,"value":316},"Which language do you plan to write the program in?",{"type":45,"value":318}," C \u002F C++, Rust, Go, Python, other?",{"type":40,"tag":198,"props":320,"children":321},{},[322,324,330],{"type":45,"value":323},"DOCA's public surface is a C ABI. Non-C consumers go through FFI \u002F language bindings (",{"type":40,"tag":65,"props":325,"children":327},{"className":326},[],[328],{"type":45,"value":329},"doca-programming-guide CAPABILITIES.md ## Capabilities and modes",{"type":45,"value":331}," and the per-library skill). The C samples are the reference even when the user's language is not C.",{"type":40,"tag":198,"props":333,"children":334},{},[335,337,342,344,349,351,355],{"type":45,"value":336},"Picks whether the agent's first-app guidance is ",{"type":40,"tag":135,"props":338,"children":339},{},[340],{"type":45,"value":341},"direct C build",{"type":45,"value":343}," or ",{"type":40,"tag":135,"props":345,"children":346},{},[347],{"type":45,"value":348},"FFI \u002F bindings against the C ABI",{"type":45,"value":350},". Does ",{"type":40,"tag":52,"props":352,"children":353},{},[354],{"type":45,"value":159},{"type":45,"value":356}," change which sample the agent points at first.",{"type":40,"tag":48,"props":358,"children":359},{},[360,362,367],{"type":45,"value":361},"The four-question gate applies when an open-ended orientation, install,\nbuild, or run recommendation depends on the user's environment. For that\nclass, ",{"type":40,"tag":52,"props":363,"children":364},{},[365],{"type":45,"value":366},"never recommend a specific install path, container tag, or runnable\nsample without first having the answers to questions 1–3",{"type":45,"value":368}," (question 4 is\nneeded for the first-app workflow but not for orientation itself). Pure\ndocumentation lookups — a URL, API reference, sample location, version\nreference, or stale-link recovery — skip this gate and route directly.\nIf the user cannot or will not answer a required question, state which\nenvironment-dependent recommendation cannot be made, give only the safe\nversion-agnostic umbrella documentation route, and do not guess the missing\nvalue. Volunteering specific commands before this is the single most common\nfailure mode for DOCA orientation.",{"type":40,"tag":48,"props":370,"children":371},{},[372],{"type":45,"value":373},"If the user has already volunteered some of the information in their first\nmessage, mark those questions answered and only ask the rest. Do not re-ask\nwhat the user has already told you.",{"type":40,"tag":74,"props":375,"children":377},{"id":376},"topic-to-where-to-look-first-routing-table",[378],{"type":45,"value":379},"Topic to \"where to look first\" routing table",{"type":40,"tag":48,"props":381,"children":382},{},[383],{"type":45,"value":384},"When the user asks something, route as follows:",{"type":40,"tag":163,"props":386,"children":387},{},[388,404],{"type":40,"tag":167,"props":389,"children":390},{},[391],{"type":40,"tag":171,"props":392,"children":393},{},[394,399],{"type":40,"tag":175,"props":395,"children":396},{},[397],{"type":45,"value":398},"User intent",{"type":40,"tag":175,"props":400,"children":401},{},[402],{"type":45,"value":403},"First place to look",{"type":40,"tag":191,"props":405,"children":406},{},[407,420,461,489,509,536,564,601,650,663,686,709,722,735,748],{"type":40,"tag":171,"props":408,"children":409},{},[410,415],{"type":40,"tag":198,"props":411,"children":412},{},[413],{"type":45,"value":414},"\"How do I install DOCA?\"",{"type":40,"tag":198,"props":416,"children":417},{},[418],{"type":45,"value":419},"Installation Guide + Downloads page (Public documentation entry points).",{"type":40,"tag":171,"props":421,"children":422},{},[423,428],{"type":40,"tag":198,"props":424,"children":425},{},[426],{"type":45,"value":427},"\"How do I start with DOCA — what's the very first thing?\"",{"type":40,"tag":198,"props":429,"children":430},{},[431,433,438,440,448,450,459],{"type":45,"value":432},"Developer Quick Start Guide ",{"type":40,"tag":135,"props":434,"children":435},{},[436],{"type":45,"value":437},"if",{"type":45,"value":439}," the user has BlueField + host hardware; otherwise ",{"type":40,"tag":60,"props":441,"children":442},{"href":222},[443],{"type":40,"tag":65,"props":444,"children":446},{"className":445},[],[447],{"type":45,"value":229},{"type":45,"value":449}," Path 0 (NGC container). Use the four questions in ",{"type":40,"tag":60,"props":451,"children":453},{"href":452},"#first-contact-discovery--the-four-questions-to-ask-before-any-drill-down",[454],{"type":40,"tag":135,"props":455,"children":456},{},[457],{"type":45,"value":458},"First-contact discovery",{"type":45,"value":460}," to pick.",{"type":40,"tag":171,"props":462,"children":463},{},[464,469],{"type":40,"tag":198,"props":465,"children":466},{},[467],{"type":45,"value":468},"\"Do I need a BlueField? A SmartNIC? Or just DOCA-Host?\"",{"type":40,"tag":198,"props":470,"children":471},{},[472,474,480,482,487],{"type":45,"value":473},"Overview page (",{"type":40,"tag":65,"props":475,"children":477},{"className":476},[],[478],{"type":45,"value":479},"doca-overview\u002Findex.html",{"type":45,"value":481},") plus the Installation Guide's ",{"type":40,"tag":135,"props":483,"children":484},{},[485],{"type":45,"value":486},"DOCA installation profiles",{"type":45,"value":488}," section. The bundle does not pick the hardware for the user — these two pages do.",{"type":40,"tag":171,"props":490,"children":491},{},[492,497],{"type":40,"tag":198,"props":493,"children":494},{},[495],{"type":45,"value":496},"\"Which package gives me library X?\"",{"type":40,"tag":198,"props":498,"children":499},{},[500,502,508],{"type":45,"value":501},"Installation Guide section on package matrix; then verify on the user's system with ",{"type":40,"tag":65,"props":503,"children":505},{"className":504},[],[506],{"type":45,"value":507},"pkg-config --list-all",{"type":45,"value":110},{"type":40,"tag":171,"props":510,"children":511},{},[512,517],{"type":40,"tag":198,"props":513,"children":514},{},[515],{"type":45,"value":516},"\"Show me a sample that uses library X.\"",{"type":40,"tag":198,"props":518,"children":519},{},[520,526,528,534],{"type":40,"tag":65,"props":521,"children":523},{"className":522},[],[524],{"type":45,"value":525},"\u002Fopt\u002Fmellanox\u002Fdoca\u002Fsamples\u002Fdoca_\u003CX>\u002F",{"type":45,"value":527}," if installed; otherwise the per-library guide on ",{"type":40,"tag":65,"props":529,"children":531},{"className":530},[],[532],{"type":45,"value":533},"docs.nvidia.com\u002Fdoca\u002Fsdk\u002F",{"type":45,"value":535}," (each library guide documents the samples shipped with it).",{"type":40,"tag":171,"props":537,"children":538},{},[539,544],{"type":40,"tag":198,"props":540,"children":541},{},[542],{"type":45,"value":543},"\"How do I build a DOCA sample?\"",{"type":40,"tag":198,"props":545,"children":546},{},[547,549,555,557,563],{"type":45,"value":548},"Library guide + the sample's own ",{"type":40,"tag":65,"props":550,"children":552},{"className":551},[],[553],{"type":45,"value":554},"meson.build",{"type":45,"value":556}," inside ",{"type":40,"tag":65,"props":558,"children":560},{"className":559},[],[561],{"type":45,"value":562},"\u002Fopt\u002Fmellanox\u002Fdoca\u002Fsamples\u002F...",{"type":45,"value":110},{"type":40,"tag":171,"props":565,"children":566},{},[567,572],{"type":40,"tag":198,"props":568,"children":569},{},[570],{"type":45,"value":571},"\"What is the API for X?\"",{"type":40,"tag":198,"props":573,"children":574},{},[575,577,583,585,591,593,599],{"type":45,"value":576},"Library guide; confirm by inspecting headers under ",{"type":40,"tag":65,"props":578,"children":580},{"className":579},[],[581],{"type":45,"value":582},"$(pkg-config --variable=includedir doca-common)",{"type":45,"value":584}," (commonly ",{"type":40,"tag":65,"props":586,"children":588},{"className":587},[],[589],{"type":45,"value":590},"\u002Fopt\u002Fmellanox\u002Fdoca\u002Finclude\u002F",{"type":45,"value":592}," on DOCA 3.3+; ",{"type":40,"tag":65,"props":594,"children":596},{"className":595},[],[597],{"type":45,"value":598},"\u002Fopt\u002Fmellanox\u002Fdoca\u002Finfrastructure\u002Finclude\u002F",{"type":45,"value":600}," on legacy \u002F split-profile installs).",{"type":40,"tag":171,"props":602,"children":603},{},[604,625],{"type":40,"tag":198,"props":605,"children":606},{},[607,609,615,617,623],{"type":45,"value":608},"\"Why does my build fail with ",{"type":40,"tag":65,"props":610,"children":612},{"className":611},[],[613],{"type":45,"value":614},"pkg-config",{"type":45,"value":616}," not finding ",{"type":40,"tag":65,"props":618,"children":620},{"className":619},[],[621],{"type":45,"value":622},"doca-...",{"type":45,"value":624},"?\"",{"type":40,"tag":198,"props":626,"children":627},{},[628,630,636,638,648],{"type":45,"value":629},"\"Layout of an installed DOCA package\" (",{"type":40,"tag":65,"props":631,"children":633},{"className":632},[],[634],{"type":45,"value":635},"PKG_CONFIG_PATH",{"type":45,"value":637},"), then load ",{"type":40,"tag":60,"props":639,"children":641},{"href":640},"..\u002Fdoca-setup\u002FSKILL.md",[642],{"type":40,"tag":65,"props":643,"children":645},{"className":644},[],[646],{"type":45,"value":647},"doca-setup",{"type":45,"value":649}," for environment preparation and debugging; do not silently mutate the environment.",{"type":40,"tag":171,"props":651,"children":652},{},[653,658],{"type":40,"tag":198,"props":654,"children":655},{},[656],{"type":45,"value":657},"\"What is the latest version \u002F what changed?\"",{"type":40,"tag":198,"props":659,"children":660},{},[661],{"type":45,"value":662},"Release Notes.",{"type":40,"tag":171,"props":664,"children":665},{},[666,671],{"type":40,"tag":198,"props":667,"children":668},{},[669],{"type":45,"value":670},"\"Is there a newer DOCA — should I upgrade \u002F downgrade \u002F roll back?\"",{"type":40,"tag":198,"props":672,"children":673},{},[674,684],{"type":40,"tag":60,"props":675,"children":677},{"href":676},"..\u002Fdoca-upgrade\u002FSKILL.md",[678],{"type":40,"tag":65,"props":679,"children":681},{"className":680},[],[682],{"type":45,"value":683},"doca-upgrade",{"type":45,"value":685}," for the detect → report → ASK → guided-upgrade discipline (never auto-upgrade); Release Notes for what the target release contains.",{"type":40,"tag":171,"props":687,"children":688},{},[689,694],{"type":40,"tag":198,"props":690,"children":691},{},[692],{"type":45,"value":693},"\"Is the component I depend on being sunset \u002F deprecated?\"",{"type":40,"tag":198,"props":695,"children":696},{},[697,699,707],{"type":45,"value":698},"Release Notes (lifecycle \u002F deprecation notices), reached via the ",{"type":40,"tag":60,"props":700,"children":701},{"href":676},[702],{"type":40,"tag":65,"props":703,"children":705},{"className":704},[],[706],{"type":45,"value":683},{"type":45,"value":708}," sunset-awareness check; do not assert deprecation from memory.",{"type":40,"tag":171,"props":710,"children":711},{},[712,717],{"type":40,"tag":198,"props":713,"children":714},{},[715],{"type":45,"value":716},"\"What does the DOCA version number mean? Is LTS still supported?\"",{"type":40,"tag":198,"props":718,"children":719},{},[720],{"type":45,"value":721},"Compatibility Policy (Public documentation entry points table).",{"type":40,"tag":171,"props":723,"children":724},{},[725,730],{"type":40,"tag":198,"props":726,"children":727},{},[728],{"type":45,"value":729},"\"How do I run DOCA on Kubernetes \u002F provision a DPU?\"",{"type":40,"tag":198,"props":731,"children":732},{},[733],{"type":45,"value":734},"DOCA Platform Framework on GitHub.",{"type":40,"tag":171,"props":736,"children":737},{},[738,743],{"type":40,"tag":198,"props":739,"children":740},{},[741],{"type":45,"value":742},"\"I have a behavior I cannot explain.\"",{"type":40,"tag":198,"props":744,"children":745},{},[746],{"type":45,"value":747},"Release Notes (known issues) first; then the DOCA Developer Forum. Never go to the forum first.",{"type":40,"tag":171,"props":749,"children":750},{},[751,756],{"type":40,"tag":198,"props":752,"children":753},{},[754],{"type":45,"value":755},"Any intent not matched above",{"type":40,"tag":198,"props":757,"children":758},{},[759,761,771],{"type":45,"value":760},"Start at the DOCA SDK index and the Libraries \u002F Services \u002F Tools umbrella page in ",{"type":40,"tag":60,"props":762,"children":764},{"href":763},"references\u002Fmap.md#public-documentation-entry-points",[765],{"type":40,"tag":65,"props":766,"children":768},{"className":767},[],[769],{"type":45,"value":770},"references\u002Fmap.md",{"type":45,"value":772},". If the authoritative route remains unresolved, say so and use the Developer Forum; do not invent a URL or component name.",{"type":40,"tag":74,"props":774,"children":776},{"id":775},"what-this-skill-deliberately-does-not-cover",[777],{"type":45,"value":778},"What this skill deliberately does not cover",{"type":40,"tag":48,"props":780,"children":781},{},[782,784,789],{"type":45,"value":783},"This file is intentionally a ",{"type":40,"tag":52,"props":785,"children":786},{},[787],{"type":45,"value":788},"map",{"type":45,"value":790},", not a tutorial. It does not contain:",{"type":40,"tag":792,"props":793,"children":794},"ul",{},[795,801,806,818],{"type":40,"tag":796,"props":797,"children":798},"li",{},[799],{"type":45,"value":800},"DOCA library tutorials (those live in the per-library guides).",{"type":40,"tag":796,"props":802,"children":803},{},[804],{"type":45,"value":805},"API reference (lives in headers and the per-library guides).",{"type":40,"tag":796,"props":807,"children":808},{},[809,811,816],{"type":45,"value":810},"Build-system deep-dives (lives in the Installation Guide and the sample\n",{"type":40,"tag":65,"props":812,"children":814},{"className":813},[],[815],{"type":45,"value":554},{"type":45,"value":817}," files).",{"type":40,"tag":796,"props":819,"children":820},{},[821],{"type":45,"value":822},"Performance tuning, driver-level setup, OFED interaction (lives in the\nInstallation Guide and library-specific guides).",{"type":40,"tag":48,"props":824,"children":825},{},[826,828,834],{"type":45,"value":827},"When the agent needs those, it should fetch the matching public document or\nread the matching installed file. As more focused skills are added, they\nshould appear in ",{"type":40,"tag":60,"props":829,"children":831},{"href":830},"..\u002F..\u002FSKILLS.md",[832],{"type":45,"value":833},"SKILLS.md",{"type":45,"value":835}," and link back here for the\n\"where to look\" lookups.",{"type":40,"tag":74,"props":837,"children":839},{"id":838},"related-skills",[840],{"type":45,"value":841},"Related skills",{"type":40,"tag":48,"props":843,"children":844},{},[845,847,852,854,859,861,867,869,877,879,884],{"type":45,"value":846},"For env preparation — install verification, build environment\n(",{"type":40,"tag":65,"props":848,"children":850},{"className":849},[],[851],{"type":45,"value":614},{"type":45,"value":853},", headers, hugepages, devlink), env-class debugging, and\nthe ",{"type":40,"tag":135,"props":855,"children":856},{},[857],{"type":45,"value":858},"I have no install yet",{"type":45,"value":860}," path with the public NGC DOCA container\n(",{"type":40,"tag":65,"props":862,"children":864},{"className":863},[],[865],{"type":45,"value":866},"nvcr.io\u002Fnvidia\u002Fdoca\u002Fdoca",{"type":45,"value":868},") as the universal Stage-1 fallback for any\nuser on macOS, Windows, or Linux without DOCA — load\n",{"type":40,"tag":60,"props":870,"children":871},{"href":640},[872],{"type":40,"tag":65,"props":873,"children":875},{"className":874},[],[876],{"type":45,"value":647},{"type":45,"value":878},". That skill stops at ",{"type":40,"tag":135,"props":880,"children":881},{},[882],{"type":45,"value":883},"\"the\ninstall is healthy and the env is ready\"",{"type":45,"value":110},{"type":40,"tag":48,"props":886,"children":887},{},[888,890,896,898,903,905,911,913,919,921,931,933,939],{"type":45,"value":889},"For general DOCA programming patterns shared across every library —\nthe canonical ",{"type":40,"tag":65,"props":891,"children":893},{"className":892},[],[894],{"type":45,"value":895},"pkg-config doca-\u003Clibrary>",{"type":45,"value":897}," build pattern (C\u002FC++ direct\nor non-C via FFI), the universal ",{"type":40,"tag":135,"props":899,"children":900},{},[901],{"type":45,"value":902},"derive a custom first app from a\nshipped sample",{"type":45,"value":904}," workflow, the universal lifecycle (",{"type":40,"tag":65,"props":906,"children":908},{"className":907},[],[909],{"type":45,"value":910},"cfg-create → init → start → use → stop → destroy",{"type":45,"value":912},"), the cross-library\n",{"type":40,"tag":65,"props":914,"children":916},{"className":915},[],[917],{"type":45,"value":918},"DOCA_ERROR_*",{"type":45,"value":920}," taxonomy, and the program-side debug order — load\n",{"type":40,"tag":60,"props":922,"children":924},{"href":923},"..\u002Fdoca-programming-guide\u002FSKILL.md",[925],{"type":40,"tag":65,"props":926,"children":928},{"className":927},[],[929],{"type":45,"value":930},"doca-programming-guide",{"type":45,"value":932},". Each\nlibrary skill extends its ",{"type":40,"tag":65,"props":934,"children":936},{"className":935},[],[937],{"type":45,"value":938},"## modify",{"type":45,"value":940}," (first-app derivation) with\nlibrary-specific overrides.",{"type":40,"tag":48,"props":942,"children":943},{},[944,946,951,953,963,965,970,972,977],{"type":45,"value":945},"For DOCA Flow internals — port and representor setup, pipe creation,\nmatch\u002Faction specifications, pipe validation before hardware programming,\nFlow counters and traces, Flow version compatibility, and debugging\n",{"type":40,"tag":65,"props":947,"children":949},{"className":948},[],[950],{"type":45,"value":918},{"type":45,"value":952}," failures from the Flow API — load\n",{"type":40,"tag":60,"props":954,"children":956},{"href":955},"..\u002Flibs\u002Fdoca-flow\u002FSKILL.md",[957],{"type":40,"tag":65,"props":958,"children":960},{"className":959},[],[961],{"type":45,"value":962},"doca-flow",{"type":45,"value":964},". That skill assumes this one is\navailable for shared documentation routing and install-layout lookups,\n",{"type":40,"tag":65,"props":966,"children":968},{"className":967},[],[969],{"type":45,"value":647},{"type":45,"value":971}," for environment preparation, and ",{"type":40,"tag":65,"props":973,"children":975},{"className":974},[],[976],{"type":45,"value":930},{"type":45,"value":978},"\nfor the cross-library programming patterns it layers on top of.",{"type":40,"tag":74,"props":980,"children":982},{"id":981},"example-questions-this-skill-answers-well",[983],{"type":45,"value":984},"Example questions this skill answers well",{"type":40,"tag":48,"props":986,"children":987},{},[988,990,999],{"type":45,"value":989},"See ",{"type":40,"tag":60,"props":991,"children":993},{"href":992},"references\u002Fmap.md#example-questions-this-skill-answers-well",[994],{"type":40,"tag":65,"props":995,"children":997},{"className":996},[],[998],{"type":45,"value":770},{"type":45,"value":110},{"type":40,"tag":74,"props":1001,"children":1003},{"id":1002},"library-and-module-specific-guides",[1004],{"type":45,"value":301},{"type":40,"tag":48,"props":1006,"children":1007},{},[1008,1009,1018],{"type":45,"value":989},{"type":40,"tag":60,"props":1010,"children":1012},{"href":1011},"references\u002Fmap.md#library--and-module-specific-guides",[1013],{"type":40,"tag":65,"props":1014,"children":1016},{"className":1015},[],[1017],{"type":45,"value":770},{"type":45,"value":110},{"type":40,"tag":74,"props":1020,"children":1022},{"id":1021},"doca-services",[1023],{"type":45,"value":1024},"DOCA services",{"type":40,"tag":48,"props":1026,"children":1027},{},[1028,1029,1038],{"type":45,"value":989},{"type":40,"tag":60,"props":1030,"children":1032},{"href":1031},"references\u002Fmap.md#doca-services",[1033],{"type":40,"tag":65,"props":1034,"children":1036},{"className":1035},[],[1037],{"type":45,"value":770},{"type":45,"value":110},{"type":40,"tag":74,"props":1040,"children":1042},{"id":1041},"doca-tools",[1043],{"type":45,"value":1044},"DOCA tools",{"type":40,"tag":48,"props":1046,"children":1047},{},[1048,1049,1058],{"type":45,"value":989},{"type":40,"tag":60,"props":1050,"children":1052},{"href":1051},"references\u002Fmap.md#doca-tools",[1053],{"type":40,"tag":65,"props":1054,"children":1056},{"className":1055},[],[1057],{"type":45,"value":770},{"type":45,"value":110},{"type":40,"tag":74,"props":1060,"children":1062},{"id":1061},"externally-productized-doca-software-not-in-this-bundle-but-here-is-where-to-route",[1063],{"type":45,"value":1064},"Externally-productized DOCA software — not in this bundle, but here is where to route",{"type":40,"tag":48,"props":1066,"children":1067},{},[1068,1069,1078],{"type":45,"value":989},{"type":40,"tag":60,"props":1070,"children":1072},{"href":1071},"references\u002Fmap.md#externally-productized-doca-software--not-in-this-bundle-but-here-is-where-to-route",[1073],{"type":40,"tag":65,"props":1074,"children":1076},{"className":1075},[],[1077],{"type":45,"value":770},{"type":45,"value":110},{"type":40,"tag":74,"props":1080,"children":1082},{"id":1081},"public-source-code-github",[1083],{"type":45,"value":1084},"Public source code: GitHub",{"type":40,"tag":48,"props":1086,"children":1087},{},[1088,1089,1098],{"type":45,"value":989},{"type":40,"tag":60,"props":1090,"children":1092},{"href":1091},"references\u002Fmap.md#public-source-code-github",[1093],{"type":40,"tag":65,"props":1094,"children":1096},{"className":1095},[],[1097],{"type":45,"value":770},{"type":45,"value":110},{"type":40,"tag":74,"props":1100,"children":1102},{"id":1101},"layout-of-an-installed-doca-package",[1103],{"type":45,"value":1104},"Layout of an installed DOCA package",{"type":40,"tag":48,"props":1106,"children":1107},{},[1108,1109,1118],{"type":45,"value":989},{"type":40,"tag":60,"props":1110,"children":1112},{"href":1111},"references\u002Fmap.md#layout-of-an-installed-doca-package",[1113],{"type":40,"tag":65,"props":1114,"children":1116},{"className":1115},[],[1117],{"type":45,"value":770},{"type":45,"value":110},{"type":40,"tag":74,"props":1120,"children":1122},{"id":1121},"where-to-find-the-version",[1123],{"type":45,"value":1124},"Where to find the version",{"type":40,"tag":48,"props":1126,"children":1127},{},[1128,1129,1138],{"type":45,"value":989},{"type":40,"tag":60,"props":1130,"children":1132},{"href":1131},"references\u002Fmap.md#where-to-find-the-version",[1133],{"type":40,"tag":65,"props":1134,"children":1136},{"className":1135},[],[1137],{"type":45,"value":770},{"type":45,"value":110},{"type":40,"tag":74,"props":1140,"children":1142},{"id":1141},"url-audit",[1143],{"type":45,"value":1144},"URL audit",{"type":40,"tag":48,"props":1146,"children":1147},{},[1148,1149,1158],{"type":45,"value":989},{"type":40,"tag":60,"props":1150,"children":1152},{"href":1151},"references\u002Fmap.md#url-audit",[1153],{"type":40,"tag":65,"props":1154,"children":1156},{"className":1155},[],[1157],{"type":45,"value":770},{"type":45,"value":110},{"type":40,"tag":74,"props":1160,"children":1162},{"id":1161},"ground-rules-for-any-agent-using-this-skill",[1163],{"type":45,"value":1164},"Ground rules for any agent using this skill",{"type":40,"tag":48,"props":1166,"children":1167},{},[1168,1169,1178],{"type":45,"value":989},{"type":40,"tag":60,"props":1170,"children":1172},{"href":1171},"references\u002Fmap.md#ground-rules-for-any-agent-using-this-skill",[1173],{"type":40,"tag":65,"props":1174,"children":1176},{"className":1175},[],[1177],{"type":45,"value":770},{"type":45,"value":110},{"type":40,"tag":74,"props":1180,"children":1182},{"id":1181},"public-documentation-entry-points",[1183],{"type":45,"value":1184},"Public documentation entry points",{"type":40,"tag":48,"props":1186,"children":1187},{},[1188,1189,1197],{"type":45,"value":989},{"type":40,"tag":60,"props":1190,"children":1191},{"href":763},[1192],{"type":40,"tag":65,"props":1193,"children":1195},{"className":1194},[],[1196],{"type":45,"value":770},{"type":45,"value":110},{"items":1199,"total":1356},[1200,1216,1234,1245,1257,1271,1284,1298,1311,1322,1336,1345],{"slug":1201,"name":1201,"fn":1202,"description":1203,"org":1204,"tags":1205,"stars":1213,"repoUrl":1214,"updatedAt":1215},"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},[1206,1207,1210],{"name":13,"slug":14,"type":15},{"name":1208,"slug":1209,"type":15},"MCP","mcp",{"name":1211,"slug":1212,"type":15},"Search","search",21777,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FNemoClaw","2026-07-20T06:00:01.461044",{"slug":1217,"name":1217,"fn":1218,"description":1219,"org":1220,"tags":1221,"stars":1231,"repoUrl":1232,"updatedAt":1233},"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},[1222,1225,1228],{"name":1223,"slug":1224,"type":15},"Containers","containers",{"name":1226,"slug":1227,"type":15},"Deployment","deployment",{"name":1229,"slug":1230,"type":15},"Python","python",17049,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FMegatron-LM","2026-07-27T06:06:11.249662",{"slug":1235,"name":1235,"fn":1236,"description":1237,"org":1238,"tags":1239,"stars":1231,"repoUrl":1232,"updatedAt":1244},"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},[1240,1243],{"name":1241,"slug":1242,"type":15},"CI\u002FCD","ci-cd",{"name":1226,"slug":1227,"type":15},"2026-07-14T05:25:59.97109",{"slug":1246,"name":1246,"fn":1247,"description":1248,"org":1249,"tags":1250,"stars":1231,"repoUrl":1232,"updatedAt":1256},"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},[1251,1252,1253],{"name":1241,"slug":1242,"type":15},{"name":1226,"slug":1227,"type":15},{"name":1254,"slug":1255,"type":15},"GitHub","github","2026-07-27T06:06:12.278222",{"slug":1258,"name":1258,"fn":1259,"description":1260,"org":1261,"tags":1262,"stars":1231,"repoUrl":1232,"updatedAt":1270},"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},[1263,1266,1267],{"name":1264,"slug":1265,"type":15},"Debugging","debugging",{"name":1254,"slug":1255,"type":15},{"name":1268,"slug":1269,"type":15},"Triage","triage","2026-07-14T05:25:57.442089",{"slug":1272,"name":1272,"fn":1273,"description":1274,"org":1275,"tags":1276,"stars":1231,"repoUrl":1232,"updatedAt":1283},"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},[1277,1280],{"name":1278,"slug":1279,"type":15},"Best Practices","best-practices",{"name":1281,"slug":1282,"type":15},"Code Analysis","code-analysis","2026-07-14T05:25:56.18433",{"slug":1285,"name":1285,"fn":1286,"description":1287,"org":1288,"tags":1289,"stars":1231,"repoUrl":1232,"updatedAt":1297},"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},[1290,1293,1296],{"name":1291,"slug":1292,"type":15},"Machine Learning","machine-learning",{"name":1294,"slug":1295,"type":15},"Migration","migration",{"name":9,"slug":8,"type":15},"2026-07-17T06:07:11.777011",{"slug":1299,"name":1299,"fn":1300,"description":1301,"org":1302,"tags":1303,"stars":1231,"repoUrl":1232,"updatedAt":1310},"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},[1304,1307],{"name":1305,"slug":1306,"type":15},"QA","qa",{"name":1308,"slug":1309,"type":15},"Testing","testing","2026-07-14T05:25:53.673039",{"slug":1312,"name":1312,"fn":1313,"description":1314,"org":1315,"tags":1316,"stars":1231,"repoUrl":1232,"updatedAt":1321},"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},[1317,1318],{"name":1226,"slug":1227,"type":15},{"name":1319,"slug":1320,"type":15},"Infrastructure","infrastructure","2026-07-14T05:25:49.362534",{"slug":1323,"name":1323,"fn":1324,"description":1325,"org":1326,"tags":1327,"stars":1231,"repoUrl":1232,"updatedAt":1335},"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},[1328,1331,1332],{"name":1329,"slug":1330,"type":15},"Code Review","code-review",{"name":1254,"slug":1255,"type":15},{"name":1333,"slug":1334,"type":15},"Pull Requests","pull-requests","2026-07-14T05:26:01.226578",{"slug":1337,"name":1337,"fn":1338,"description":1339,"org":1340,"tags":1341,"stars":1231,"repoUrl":1232,"updatedAt":1344},"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},[1342,1343],{"name":1305,"slug":1306,"type":15},{"name":1308,"slug":1309,"type":15},"2026-07-14T05:25:54.928983",{"slug":1346,"name":1346,"fn":1347,"description":1348,"org":1349,"tags":1350,"stars":1231,"repoUrl":1232,"updatedAt":1355},"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},[1351,1354],{"name":1352,"slug":1353,"type":15},"Automation","automation",{"name":1241,"slug":1242,"type":15},"2026-07-30T05:29:03.275638",496,{"items":1358,"total":1454},[1359,1376,1386,1400,1410,1425,1440],{"slug":1360,"name":1360,"fn":1361,"description":1362,"org":1363,"tags":1364,"stars":20,"repoUrl":21,"updatedAt":1375},"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},[1365,1368,1371,1372],{"name":1366,"slug":1367,"type":15},"Data Analysis","data-analysis",{"name":1369,"slug":1370,"type":15},"Data Engineering","data-engineering",{"name":9,"slug":8,"type":15},{"name":1373,"slug":1374,"type":15},"Performance","performance","2026-07-14T05:28:43.176466",{"slug":1377,"name":1377,"fn":1378,"description":1379,"org":1380,"tags":1381,"stars":20,"repoUrl":21,"updatedAt":1385},"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},[1382,1383,1384],{"name":1226,"slug":1227,"type":15},{"name":1319,"slug":1320,"type":15},{"name":9,"slug":8,"type":15},"2026-07-14T05:29:06.667109",{"slug":1387,"name":1387,"fn":1388,"description":1389,"org":1390,"tags":1391,"stars":20,"repoUrl":21,"updatedAt":1399},"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},[1392,1395,1396],{"name":1393,"slug":1394,"type":15},"Agents","agents",{"name":9,"slug":8,"type":15},{"name":1397,"slug":1398,"type":15},"Research","research","2026-07-14T05:28:06.816956",{"slug":1401,"name":1401,"fn":1402,"description":1403,"org":1404,"tags":1405,"stars":20,"repoUrl":21,"updatedAt":1409},"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},[1406,1407,1408],{"name":1366,"slug":1367,"type":15},{"name":9,"slug":8,"type":15},{"name":1308,"slug":1309,"type":15},"2026-07-17T05:29:03.913266",{"slug":1411,"name":1411,"fn":1412,"description":1413,"org":1414,"tags":1415,"stars":20,"repoUrl":21,"updatedAt":1424},"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},[1416,1417,1420,1421],{"name":1352,"slug":1353,"type":15},{"name":1418,"slug":1419,"type":15},"Imaging","imaging",{"name":9,"slug":8,"type":15},{"name":1422,"slug":1423,"type":15},"Video","video","2026-07-17T05:28:53.905004",{"slug":1426,"name":1426,"fn":1427,"description":1428,"org":1429,"tags":1430,"stars":20,"repoUrl":21,"updatedAt":1439},"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},[1431,1432,1435,1436],{"name":1226,"slug":1227,"type":15},{"name":1433,"slug":1434,"type":15},"Docker","docker",{"name":9,"slug":8,"type":15},{"name":1437,"slug":1438,"type":15},"Operations","operations","2026-07-17T05:28:56.913999",{"slug":1441,"name":1441,"fn":1442,"description":1443,"org":1444,"tags":1445,"stars":20,"repoUrl":21,"updatedAt":1453},"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},[1446,1447,1450],{"name":9,"slug":8,"type":15},{"name":1448,"slug":1449,"type":15},"Quantum Computing","quantum-computing",{"name":1451,"slug":1452,"type":15},"Simulation","simulation","2026-07-14T05:26:58.898253",305]