[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill-nvidia-vss-ask-video":3,"mdc--rmkhm9-key":31,"related-org-nvidia-vss-ask-video":1420,"related-repo-nvidia-vss-ask-video":1581},{"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},"vss-ask-video","analyze video clips with visual questions","Use this skill to ask the VSS agent's video_understanding tool a fresh visual question about a recorded clip. Not for prior tool output, search hits, or metadata-answerable questions.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},"nvidia","NVIDIA","https:\u002F\u002Fpexgzepcugksgbtrxkhf.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Forg-logos\u002Fnvidia.png",[12,14,17],{"name":9,"slug":8,"type":13},"tag",{"name":15,"slug":16,"type":13},"Computer Vision","computer-vision",{"name":18,"slug":19,"type":13},"Video","video",2473,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Fskills","2026-07-14T05:29:48.184888","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\u002Fvss-ask-video","---\nname: vss-ask-video\ndescription: Use this skill to ask the VSS agent's video_understanding tool a fresh visual question about a recorded clip. Not for prior tool output, search hits, or metadata-answerable questions.\nlicense: Apache-2.0\nmetadata:\n  version: \"3.2.0\"\n  github-url: \"https:\u002F\u002Fgithub.com\u002FNVIDIA-AI-Blueprints\u002Fvideo-search-and-summarization\"\n  tags: \"nvidia blueprint operational\"\n---\n\n# Video QnA using VLM through VSS Agent\n\nUse this skill when you need details about the video which requires VLM to look at the video frames — for example the agent has **no** usable prior answer and needs a **fresh look at the pixels** for a specific clip.\n\n---\n\n## When to Use\n\n- The user asks **what happens in the video**, what **objects \u002F people \u002F actions** appear, **colors**, **timing**, **safety**, or other **visual facts** that require watching the clip.\n- The user asks for **details** that **cannot be answered** from existing messages, summaries, Elasticsearch\u002FMCP results, or filenames alone—you need **model inference on the video**.\n- Follow-up questions about **content details** after a coarse summary or after report generation.\n\nDo **not** use this skill when a **database \u002F MCP \u002F prior tool output** already answers the question, unless the user explicitly wants **verification** against the video.\n\n---\n\n## Deployment prerequisite\n\nThis skill requires a VSS profile that serves the `video_understanding` tool — typically **base** (recommended) or **lvs**. Before any request:\n\n1. Probe the VSS agent:\n   ```bash\n   curl -sf --max-time 5 \"http:\u002F\u002F${HOST_IP}:8000\u002Fdocs\" >\u002Fdev\u002Fnull\n   ```\n\n2. **If the probe fails**, ask the user:\n   > *\"No VSS profile is running on `$HOST_IP`. Shall I deploy `base` (recommended for per-clip VLM QnA) using the `\u002Fvss-deploy-profile` skill? If you prefer `lvs`, say so.\"*\n\n   - If yes → hand off to `\u002Fvss-deploy-profile -p base` (or `-p lvs` if the user prefers). Return here once it succeeds.\n   - If no → stop.\n\n3. If the probe passes, proceed.\n\n---\n\n## Sensor prerequisite\n\n**You MUST list VST sensors before any `\u002Fgenerate` call.** This is required even when the user names the sensor explicitly, even when the user asserts the video is already uploaded, and even when a previous turn appeared to use the same video. Do not skip this step.\n\n1. List sensors:\n   ```bash\n   curl -sf --max-time 5 \"http:\u002F\u002F${HOST_IP}:30888\u002Fvst\u002Fapi\u002Fv1\u002Fsensor\u002Flist\" | jq '.[].name'\n   ```\n\n2. Compare the returned `name` values against the user-supplied `\u003Csensor-id>` (or **filename stem**, e.g. `warehouse_safety_0001`).\n\n3. **If a matching sensor is present** → proceed to the Agent workflow below.\n\n4. **If no matching sensor is present** — upload the video first, then re-list to confirm the new sensor appears:\n   ```bash\n   # filename: must not contain whitespace\n   # timestamp: ISO 8601 UTC — default 2025-01-01T00:00:00.000Z if user did not specify\n   curl -s -X PUT \"http:\u002F\u002F${HOST_IP}:30888\u002Fvst\u002Fapi\u002Fv1\u002Fstorage\u002Ffile\u002F\u003Cfilename>?timestamp=\u003Ctimestamp>\" \\\n     -H \"Content-Type: application\u002Foctet-stream\" \\\n     -H \"Content-Length: \u003Cfile_size_in_bytes>\" \\\n     --upload-file \u002Fpath\u002Fto\u002F\u003Cfilename> | jq .\n   ```\n   See `\u002Fvss-manage-video-io-storage` for full upload semantics (v1 vs v2, conflict handling, delete flow). In interactive runs, confirm with the user before uploading. **Never** issue an unconditional PUT without first running the sensor-list check above — that is exactly the failure mode this prerequisite exists to prevent.\n\n---\n\n## Agent workflow\n\nThe Sensor prerequisite above must have already confirmed (or made) the sensor exist on VST. Then:\n\n1. **Clip** — Identify **sensor id**, **filename**, or **URL** for one video segment. If ambiguous, ask the user.\n2. Call vss agent with the sensor id and ask for it to call video_understanding tool to answer the user's question.\n3. Return the vss agent's answer back to the user.\n\n\n## Query VSS agent (`\u002Fgenerate`)\n\n```bash\n# Set from deployment (compose \u002F .env \u002F host where vss-agent listens)\nexport VSS_AGENT_BASE_URL=\"http:\u002F\u002Flocalhost:8000\"\n\ncurl -s -X POST \"${VSS_AGENT_BASE_URL}\u002Fgenerate\" \\\n  -H \"Content-Type: application\u002Fjson\" \\\n  -d '{\"input_message\": \"Call video_understanding tool to answer the following question about \u003Csensor-id>: \u003Cuser query>\"}' | jq .\n```\n\n### Response contract and extraction\n\n`\u002Fgenerate` returns a JSON object with the assistant output in `value`, for example:\n\n```json\n{\"value\":\"\u003Cagent-think>\u003Cagent-think-step ...>...\u003C\u002Fagent-think-step>\u003C\u002Fagent-think>\\n\\n\u003Cfinal answer>\\n\\n\"}\n```\n\nThere is no separate clean-answer field. The consumable answer is the text in `.value` after removing any `\u003Cagent-think>...\u003C\u002Fagent-think>` block.\n\nRequired handling for this skill (and any downstream caller):\n\n1. Read `.value` from the JSON response.\n2. Strip `\u003Cagent-think>...\u003C\u002Fagent-think>` sections wherever they appear.\n3. Return only the remaining final-answer text to the user.\n\nExample extraction:\n\n```bash\ncurl -s -X POST \"${VSS_AGENT_BASE_URL}\u002Fgenerate\" \\\n  -H \"Content-Type: application\u002Fjson\" \\\n  -d '{\"input_message\":\"Call video_understanding tool to answer the following question about \u003Csensor-id>: \u003Cuser query>\"}' \\\n| jq -r '.value' \\\n| python3 -c 'import re,sys; t=sys.stdin.read(); t=re.sub(r\"\u003Cagent-think>.*?\u003C\u002Fagent-think>\\s*\", \"\", t, flags=re.S); print(t.strip())'\n```\n\n---\n\n## Cross-Reference\n\n- **vss-manage-video-io-storage** — VST storage\u002Freplay URLs so **`VIDEO_URL`** is valid for the VLM.\n- **vss-generate-video-report** — timestamped **reports** via **Mode A (direct VLM)** or **Mode B (video-analytics incidents)**; this skill is **VSS-agent `\u002Fgenerate`** for ad-hoc **video Q&A**.\n",{"data":32,"body":37},{"name":4,"description":6,"license":23,"metadata":33},{"version":34,"github-url":35,"tags":36},"3.2.0","https:\u002F\u002Fgithub.com\u002FNVIDIA-AI-Blueprints\u002Fvideo-search-and-summarization","nvidia blueprint operational",{"type":38,"children":39},"root",[40,49,70,74,81,170,196,199,205,233,416,419,425,443,787,790,796,801,844,857,1018,1025,1043,1109,1130,1135,1167,1172,1332,1335,1341,1414],{"type":41,"tag":42,"props":43,"children":45},"element","h1",{"id":44},"video-qna-using-vlm-through-vss-agent",[46],{"type":47,"value":48},"text","Video QnA using VLM through VSS Agent",{"type":41,"tag":50,"props":51,"children":52},"p",{},[53,55,61,63,68],{"type":47,"value":54},"Use this skill when you need details about the video which requires VLM to look at the video frames — for example the agent has ",{"type":41,"tag":56,"props":57,"children":58},"strong",{},[59],{"type":47,"value":60},"no",{"type":47,"value":62}," usable prior answer and needs a ",{"type":41,"tag":56,"props":64,"children":65},{},[66],{"type":47,"value":67},"fresh look at the pixels",{"type":47,"value":69}," for a specific clip.",{"type":41,"tag":71,"props":72,"children":73},"hr",{},[],{"type":41,"tag":75,"props":76,"children":78},"h2",{"id":77},"when-to-use",[79],{"type":47,"value":80},"When to Use",{"type":41,"tag":82,"props":83,"children":84},"ul",{},[85,132,158],{"type":41,"tag":86,"props":87,"children":88},"li",{},[89,91,96,98,103,105,110,112,117,118,123,125,130],{"type":47,"value":90},"The user asks ",{"type":41,"tag":56,"props":92,"children":93},{},[94],{"type":47,"value":95},"what happens in the video",{"type":47,"value":97},", what ",{"type":41,"tag":56,"props":99,"children":100},{},[101],{"type":47,"value":102},"objects \u002F people \u002F actions",{"type":47,"value":104}," appear, ",{"type":41,"tag":56,"props":106,"children":107},{},[108],{"type":47,"value":109},"colors",{"type":47,"value":111},", ",{"type":41,"tag":56,"props":113,"children":114},{},[115],{"type":47,"value":116},"timing",{"type":47,"value":111},{"type":41,"tag":56,"props":119,"children":120},{},[121],{"type":47,"value":122},"safety",{"type":47,"value":124},", or other ",{"type":41,"tag":56,"props":126,"children":127},{},[128],{"type":47,"value":129},"visual facts",{"type":47,"value":131}," that require watching the clip.",{"type":41,"tag":86,"props":133,"children":134},{},[135,137,142,144,149,151,156],{"type":47,"value":136},"The user asks for ",{"type":41,"tag":56,"props":138,"children":139},{},[140],{"type":47,"value":141},"details",{"type":47,"value":143}," that ",{"type":41,"tag":56,"props":145,"children":146},{},[147],{"type":47,"value":148},"cannot be answered",{"type":47,"value":150}," from existing messages, summaries, Elasticsearch\u002FMCP results, or filenames alone—you need ",{"type":41,"tag":56,"props":152,"children":153},{},[154],{"type":47,"value":155},"model inference on the video",{"type":47,"value":157},".",{"type":41,"tag":86,"props":159,"children":160},{},[161,163,168],{"type":47,"value":162},"Follow-up questions about ",{"type":41,"tag":56,"props":164,"children":165},{},[166],{"type":47,"value":167},"content details",{"type":47,"value":169}," after a coarse summary or after report generation.",{"type":41,"tag":50,"props":171,"children":172},{},[173,175,180,182,187,189,194],{"type":47,"value":174},"Do ",{"type":41,"tag":56,"props":176,"children":177},{},[178],{"type":47,"value":179},"not",{"type":47,"value":181}," use this skill when a ",{"type":41,"tag":56,"props":183,"children":184},{},[185],{"type":47,"value":186},"database \u002F MCP \u002F prior tool output",{"type":47,"value":188}," already answers the question, unless the user explicitly wants ",{"type":41,"tag":56,"props":190,"children":191},{},[192],{"type":47,"value":193},"verification",{"type":47,"value":195}," against the video.",{"type":41,"tag":71,"props":197,"children":198},{},[],{"type":41,"tag":75,"props":200,"children":202},{"id":201},"deployment-prerequisite",[203],{"type":47,"value":204},"Deployment prerequisite",{"type":41,"tag":50,"props":206,"children":207},{},[208,210,217,219,224,226,231],{"type":47,"value":209},"This skill requires a VSS profile that serves the ",{"type":41,"tag":211,"props":212,"children":214},"code",{"className":213},[],[215],{"type":47,"value":216},"video_understanding",{"type":47,"value":218}," tool — typically ",{"type":41,"tag":56,"props":220,"children":221},{},[222],{"type":47,"value":223},"base",{"type":47,"value":225}," (recommended) or ",{"type":41,"tag":56,"props":227,"children":228},{},[229],{"type":47,"value":230},"lvs",{"type":47,"value":232},". Before any request:",{"type":41,"tag":234,"props":235,"children":236},"ol",{},[237,329,411],{"type":41,"tag":86,"props":238,"children":239},{},[240,242],{"type":47,"value":241},"Probe the VSS agent:",{"type":41,"tag":243,"props":244,"children":249},"pre",{"className":245,"code":246,"language":247,"meta":248,"style":248},"language-bash shiki shiki-themes material-theme-lighter material-theme material-theme-palenight","curl -sf --max-time 5 \"http:\u002F\u002F${HOST_IP}:8000\u002Fdocs\" >\u002Fdev\u002Fnull\n","bash","",[250],{"type":41,"tag":211,"props":251,"children":252},{"__ignoreMap":248},[253],{"type":41,"tag":254,"props":255,"children":258},"span",{"class":256,"line":257},"line",1,[259,265,271,276,282,288,293,298,304,309,314,319,324],{"type":41,"tag":254,"props":260,"children":262},{"style":261},"--shiki-light:#E2931D;--shiki-default:#FFCB6B;--shiki-dark:#FFCB6B",[263],{"type":47,"value":264},"curl",{"type":41,"tag":254,"props":266,"children":268},{"style":267},"--shiki-light:#91B859;--shiki-default:#C3E88D;--shiki-dark:#C3E88D",[269],{"type":47,"value":270}," -sf",{"type":41,"tag":254,"props":272,"children":273},{"style":267},[274],{"type":47,"value":275}," --max-time",{"type":41,"tag":254,"props":277,"children":279},{"style":278},"--shiki-light:#F76D47;--shiki-default:#F78C6C;--shiki-dark:#F78C6C",[280],{"type":47,"value":281}," 5",{"type":41,"tag":254,"props":283,"children":285},{"style":284},"--shiki-light:#39ADB5;--shiki-default:#89DDFF;--shiki-dark:#89DDFF",[286],{"type":47,"value":287}," \"",{"type":41,"tag":254,"props":289,"children":290},{"style":267},[291],{"type":47,"value":292},"http:\u002F\u002F",{"type":41,"tag":254,"props":294,"children":295},{"style":284},[296],{"type":47,"value":297},"${",{"type":41,"tag":254,"props":299,"children":301},{"style":300},"--shiki-light:#90A4AE;--shiki-default:#EEFFFF;--shiki-dark:#BABED8",[302],{"type":47,"value":303},"HOST_IP",{"type":41,"tag":254,"props":305,"children":306},{"style":284},[307],{"type":47,"value":308},"}",{"type":41,"tag":254,"props":310,"children":311},{"style":267},[312],{"type":47,"value":313},":8000\u002Fdocs",{"type":41,"tag":254,"props":315,"children":316},{"style":284},[317],{"type":47,"value":318},"\"",{"type":41,"tag":254,"props":320,"children":321},{"style":284},[322],{"type":47,"value":323}," >",{"type":41,"tag":254,"props":325,"children":326},{"style":267},[327],{"type":47,"value":328},"\u002Fdev\u002Fnull\n",{"type":41,"tag":86,"props":330,"children":331},{},[332,337,339,382],{"type":41,"tag":56,"props":333,"children":334},{},[335],{"type":47,"value":336},"If the probe fails",{"type":47,"value":338},", ask the user:",{"type":41,"tag":340,"props":341,"children":342},"blockquote",{},[343],{"type":41,"tag":50,"props":344,"children":345},{},[346],{"type":41,"tag":347,"props":348,"children":349},"em",{},[350,352,358,360,365,367,373,375,380],{"type":47,"value":351},"\"No VSS profile is running on ",{"type":41,"tag":211,"props":353,"children":355},{"className":354},[],[356],{"type":47,"value":357},"$HOST_IP",{"type":47,"value":359},". Shall I deploy ",{"type":41,"tag":211,"props":361,"children":363},{"className":362},[],[364],{"type":47,"value":223},{"type":47,"value":366}," (recommended for per-clip VLM QnA) using the ",{"type":41,"tag":211,"props":368,"children":370},{"className":369},[],[371],{"type":47,"value":372},"\u002Fvss-deploy-profile",{"type":47,"value":374}," skill? If you prefer ",{"type":41,"tag":211,"props":376,"children":378},{"className":377},[],[379],{"type":47,"value":230},{"type":47,"value":381},", say so.\"",{"type":41,"tag":82,"props":383,"children":384},{},[385,406],{"type":41,"tag":86,"props":386,"children":387},{},[388,390,396,398,404],{"type":47,"value":389},"If yes → hand off to ",{"type":41,"tag":211,"props":391,"children":393},{"className":392},[],[394],{"type":47,"value":395},"\u002Fvss-deploy-profile -p base",{"type":47,"value":397}," (or ",{"type":41,"tag":211,"props":399,"children":401},{"className":400},[],[402],{"type":47,"value":403},"-p lvs",{"type":47,"value":405}," if the user prefers). Return here once it succeeds.",{"type":41,"tag":86,"props":407,"children":408},{},[409],{"type":47,"value":410},"If no → stop.",{"type":41,"tag":86,"props":412,"children":413},{},[414],{"type":47,"value":415},"If the probe passes, proceed.",{"type":41,"tag":71,"props":417,"children":418},{},[],{"type":41,"tag":75,"props":420,"children":422},{"id":421},"sensor-prerequisite",[423],{"type":47,"value":424},"Sensor prerequisite",{"type":41,"tag":50,"props":426,"children":427},{},[428,441],{"type":41,"tag":56,"props":429,"children":430},{},[431,433,439],{"type":47,"value":432},"You MUST list VST sensors before any ",{"type":41,"tag":211,"props":434,"children":436},{"className":435},[],[437],{"type":47,"value":438},"\u002Fgenerate",{"type":47,"value":440}," call.",{"type":47,"value":442}," This is required even when the user names the sensor explicitly, even when the user asserts the video is already uploaded, and even when a previous turn appeared to use the same video. Do not skip this step.",{"type":41,"tag":234,"props":444,"children":445},{},[446,531,566,576],{"type":41,"tag":86,"props":447,"children":448},{},[449,451],{"type":47,"value":450},"List sensors:",{"type":41,"tag":243,"props":452,"children":454},{"className":245,"code":453,"language":247,"meta":248,"style":248},"curl -sf --max-time 5 \"http:\u002F\u002F${HOST_IP}:30888\u002Fvst\u002Fapi\u002Fv1\u002Fsensor\u002Flist\" | jq '.[].name'\n",[455],{"type":41,"tag":211,"props":456,"children":457},{"__ignoreMap":248},[458],{"type":41,"tag":254,"props":459,"children":460},{"class":256,"line":257},[461,465,469,473,477,481,485,489,493,497,502,506,511,516,521,526],{"type":41,"tag":254,"props":462,"children":463},{"style":261},[464],{"type":47,"value":264},{"type":41,"tag":254,"props":466,"children":467},{"style":267},[468],{"type":47,"value":270},{"type":41,"tag":254,"props":470,"children":471},{"style":267},[472],{"type":47,"value":275},{"type":41,"tag":254,"props":474,"children":475},{"style":278},[476],{"type":47,"value":281},{"type":41,"tag":254,"props":478,"children":479},{"style":284},[480],{"type":47,"value":287},{"type":41,"tag":254,"props":482,"children":483},{"style":267},[484],{"type":47,"value":292},{"type":41,"tag":254,"props":486,"children":487},{"style":284},[488],{"type":47,"value":297},{"type":41,"tag":254,"props":490,"children":491},{"style":300},[492],{"type":47,"value":303},{"type":41,"tag":254,"props":494,"children":495},{"style":284},[496],{"type":47,"value":308},{"type":41,"tag":254,"props":498,"children":499},{"style":267},[500],{"type":47,"value":501},":30888\u002Fvst\u002Fapi\u002Fv1\u002Fsensor\u002Flist",{"type":41,"tag":254,"props":503,"children":504},{"style":284},[505],{"type":47,"value":318},{"type":41,"tag":254,"props":507,"children":508},{"style":284},[509],{"type":47,"value":510}," |",{"type":41,"tag":254,"props":512,"children":513},{"style":261},[514],{"type":47,"value":515}," jq",{"type":41,"tag":254,"props":517,"children":518},{"style":284},[519],{"type":47,"value":520}," '",{"type":41,"tag":254,"props":522,"children":523},{"style":267},[524],{"type":47,"value":525},".[].name",{"type":41,"tag":254,"props":527,"children":528},{"style":284},[529],{"type":47,"value":530},"'\n",{"type":41,"tag":86,"props":532,"children":533},{},[534,536,542,544,550,551,556,558,564],{"type":47,"value":535},"Compare the returned ",{"type":41,"tag":211,"props":537,"children":539},{"className":538},[],[540],{"type":47,"value":541},"name",{"type":47,"value":543}," values against the user-supplied ",{"type":41,"tag":211,"props":545,"children":547},{"className":546},[],[548],{"type":47,"value":549},"\u003Csensor-id>",{"type":47,"value":397},{"type":41,"tag":56,"props":552,"children":553},{},[554],{"type":47,"value":555},"filename stem",{"type":47,"value":557},", e.g. ",{"type":41,"tag":211,"props":559,"children":561},{"className":560},[],[562],{"type":47,"value":563},"warehouse_safety_0001",{"type":47,"value":565},").",{"type":41,"tag":86,"props":567,"children":568},{},[569,574],{"type":41,"tag":56,"props":570,"children":571},{},[572],{"type":47,"value":573},"If a matching sensor is present",{"type":47,"value":575}," → proceed to the Agent workflow below.",{"type":41,"tag":86,"props":577,"children":578},{},[579,584,586,766,770,772,778,780,785],{"type":41,"tag":56,"props":580,"children":581},{},[582],{"type":47,"value":583},"If no matching sensor is present",{"type":47,"value":585}," — upload the video first, then re-list to confirm the new sensor appears:",{"type":41,"tag":243,"props":587,"children":589},{"className":245,"code":588,"language":247,"meta":248,"style":248},"# filename: must not contain whitespace\n# timestamp: ISO 8601 UTC — default 2025-01-01T00:00:00.000Z if user did not specify\ncurl -s -X PUT \"http:\u002F\u002F${HOST_IP}:30888\u002Fvst\u002Fapi\u002Fv1\u002Fstorage\u002Ffile\u002F\u003Cfilename>?timestamp=\u003Ctimestamp>\" \\\n  -H \"Content-Type: application\u002Foctet-stream\" \\\n  -H \"Content-Length: \u003Cfile_size_in_bytes>\" \\\n  --upload-file \u002Fpath\u002Fto\u002F\u003Cfilename> | jq .\n",[590],{"type":41,"tag":211,"props":591,"children":592},{"__ignoreMap":248},[593,602,611,668,694,719],{"type":41,"tag":254,"props":594,"children":595},{"class":256,"line":257},[596],{"type":41,"tag":254,"props":597,"children":599},{"style":598},"--shiki-light:#90A4AE;--shiki-light-font-style:italic;--shiki-default:#546E7A;--shiki-default-font-style:italic;--shiki-dark:#676E95;--shiki-dark-font-style:italic",[600],{"type":47,"value":601},"# filename: must not contain whitespace\n",{"type":41,"tag":254,"props":603,"children":605},{"class":256,"line":604},2,[606],{"type":41,"tag":254,"props":607,"children":608},{"style":598},[609],{"type":47,"value":610},"# timestamp: ISO 8601 UTC — default 2025-01-01T00:00:00.000Z if user did not specify\n",{"type":41,"tag":254,"props":612,"children":614},{"class":256,"line":613},3,[615,619,624,629,634,638,642,646,650,654,659,663],{"type":41,"tag":254,"props":616,"children":617},{"style":261},[618],{"type":47,"value":264},{"type":41,"tag":254,"props":620,"children":621},{"style":267},[622],{"type":47,"value":623}," -s",{"type":41,"tag":254,"props":625,"children":626},{"style":267},[627],{"type":47,"value":628}," -X",{"type":41,"tag":254,"props":630,"children":631},{"style":267},[632],{"type":47,"value":633}," PUT",{"type":41,"tag":254,"props":635,"children":636},{"style":284},[637],{"type":47,"value":287},{"type":41,"tag":254,"props":639,"children":640},{"style":267},[641],{"type":47,"value":292},{"type":41,"tag":254,"props":643,"children":644},{"style":284},[645],{"type":47,"value":297},{"type":41,"tag":254,"props":647,"children":648},{"style":300},[649],{"type":47,"value":303},{"type":41,"tag":254,"props":651,"children":652},{"style":284},[653],{"type":47,"value":308},{"type":41,"tag":254,"props":655,"children":656},{"style":267},[657],{"type":47,"value":658},":30888\u002Fvst\u002Fapi\u002Fv1\u002Fstorage\u002Ffile\u002F\u003Cfilename>?timestamp=\u003Ctimestamp>",{"type":41,"tag":254,"props":660,"children":661},{"style":284},[662],{"type":47,"value":318},{"type":41,"tag":254,"props":664,"children":665},{"style":300},[666],{"type":47,"value":667}," \\\n",{"type":41,"tag":254,"props":669,"children":671},{"class":256,"line":670},4,[672,677,681,686,690],{"type":41,"tag":254,"props":673,"children":674},{"style":267},[675],{"type":47,"value":676},"  -H",{"type":41,"tag":254,"props":678,"children":679},{"style":284},[680],{"type":47,"value":287},{"type":41,"tag":254,"props":682,"children":683},{"style":267},[684],{"type":47,"value":685},"Content-Type: application\u002Foctet-stream",{"type":41,"tag":254,"props":687,"children":688},{"style":284},[689],{"type":47,"value":318},{"type":41,"tag":254,"props":691,"children":692},{"style":300},[693],{"type":47,"value":667},{"type":41,"tag":254,"props":695,"children":697},{"class":256,"line":696},5,[698,702,706,711,715],{"type":41,"tag":254,"props":699,"children":700},{"style":267},[701],{"type":47,"value":676},{"type":41,"tag":254,"props":703,"children":704},{"style":284},[705],{"type":47,"value":287},{"type":41,"tag":254,"props":707,"children":708},{"style":267},[709],{"type":47,"value":710},"Content-Length: \u003Cfile_size_in_bytes>",{"type":41,"tag":254,"props":712,"children":713},{"style":284},[714],{"type":47,"value":318},{"type":41,"tag":254,"props":716,"children":717},{"style":300},[718],{"type":47,"value":667},{"type":41,"tag":254,"props":720,"children":722},{"class":256,"line":721},6,[723,728,733,738,743,748,753,757,761],{"type":41,"tag":254,"props":724,"children":725},{"style":267},[726],{"type":47,"value":727},"  --upload-file",{"type":41,"tag":254,"props":729,"children":730},{"style":267},[731],{"type":47,"value":732}," \u002Fpath\u002Fto\u002F",{"type":41,"tag":254,"props":734,"children":735},{"style":284},[736],{"type":47,"value":737},"\u003C",{"type":41,"tag":254,"props":739,"children":740},{"style":267},[741],{"type":47,"value":742},"filenam",{"type":41,"tag":254,"props":744,"children":745},{"style":300},[746],{"type":47,"value":747},"e",{"type":41,"tag":254,"props":749,"children":750},{"style":284},[751],{"type":47,"value":752},">",{"type":41,"tag":254,"props":754,"children":755},{"style":284},[756],{"type":47,"value":510},{"type":41,"tag":254,"props":758,"children":759},{"style":261},[760],{"type":47,"value":515},{"type":41,"tag":254,"props":762,"children":763},{"style":267},[764],{"type":47,"value":765}," .\n",{"type":41,"tag":767,"props":768,"children":769},"br",{},[],{"type":47,"value":771},"See ",{"type":41,"tag":211,"props":773,"children":775},{"className":774},[],[776],{"type":47,"value":777},"\u002Fvss-manage-video-io-storage",{"type":47,"value":779}," for full upload semantics (v1 vs v2, conflict handling, delete flow). In interactive runs, confirm with the user before uploading. ",{"type":41,"tag":56,"props":781,"children":782},{},[783],{"type":47,"value":784},"Never",{"type":47,"value":786}," issue an unconditional PUT without first running the sensor-list check above — that is exactly the failure mode this prerequisite exists to prevent.",{"type":41,"tag":71,"props":788,"children":789},{},[],{"type":41,"tag":75,"props":791,"children":793},{"id":792},"agent-workflow",[794],{"type":47,"value":795},"Agent workflow",{"type":41,"tag":50,"props":797,"children":798},{},[799],{"type":47,"value":800},"The Sensor prerequisite above must have already confirmed (or made) the sensor exist on VST. Then:",{"type":41,"tag":234,"props":802,"children":803},{},[804,834,839],{"type":41,"tag":86,"props":805,"children":806},{},[807,812,814,819,820,825,827,832],{"type":41,"tag":56,"props":808,"children":809},{},[810],{"type":47,"value":811},"Clip",{"type":47,"value":813}," — Identify ",{"type":41,"tag":56,"props":815,"children":816},{},[817],{"type":47,"value":818},"sensor id",{"type":47,"value":111},{"type":41,"tag":56,"props":821,"children":822},{},[823],{"type":47,"value":824},"filename",{"type":47,"value":826},", or ",{"type":41,"tag":56,"props":828,"children":829},{},[830],{"type":47,"value":831},"URL",{"type":47,"value":833}," for one video segment. If ambiguous, ask the user.",{"type":41,"tag":86,"props":835,"children":836},{},[837],{"type":47,"value":838},"Call vss agent with the sensor id and ask for it to call video_understanding tool to answer the user's question.",{"type":41,"tag":86,"props":840,"children":841},{},[842],{"type":47,"value":843},"Return the vss agent's answer back to the user.",{"type":41,"tag":75,"props":845,"children":847},{"id":846},"query-vss-agent-generate",[848,850,855],{"type":47,"value":849},"Query VSS agent (",{"type":41,"tag":211,"props":851,"children":853},{"className":852},[],[854],{"type":47,"value":438},{"type":47,"value":856},")",{"type":41,"tag":243,"props":858,"children":860},{"className":245,"code":859,"language":247,"meta":248,"style":248},"# Set from deployment (compose \u002F .env \u002F host where vss-agent listens)\nexport VSS_AGENT_BASE_URL=\"http:\u002F\u002Flocalhost:8000\"\n\ncurl -s -X POST \"${VSS_AGENT_BASE_URL}\u002Fgenerate\" \\\n  -H \"Content-Type: application\u002Fjson\" \\\n  -d '{\"input_message\": \"Call video_understanding tool to answer the following question about \u003Csensor-id>: \u003Cuser query>\"}' | jq .\n",[861],{"type":41,"tag":211,"props":862,"children":863},{"__ignoreMap":248},[864,872,905,914,960,984],{"type":41,"tag":254,"props":865,"children":866},{"class":256,"line":257},[867],{"type":41,"tag":254,"props":868,"children":869},{"style":598},[870],{"type":47,"value":871},"# Set from deployment (compose \u002F .env \u002F host where vss-agent listens)\n",{"type":41,"tag":254,"props":873,"children":874},{"class":256,"line":604},[875,881,886,891,895,900],{"type":41,"tag":254,"props":876,"children":878},{"style":877},"--shiki-light:#9C3EDA;--shiki-default:#C792EA;--shiki-dark:#C792EA",[879],{"type":47,"value":880},"export",{"type":41,"tag":254,"props":882,"children":883},{"style":300},[884],{"type":47,"value":885}," VSS_AGENT_BASE_URL",{"type":41,"tag":254,"props":887,"children":888},{"style":284},[889],{"type":47,"value":890},"=",{"type":41,"tag":254,"props":892,"children":893},{"style":284},[894],{"type":47,"value":318},{"type":41,"tag":254,"props":896,"children":897},{"style":267},[898],{"type":47,"value":899},"http:\u002F\u002Flocalhost:8000",{"type":41,"tag":254,"props":901,"children":902},{"style":284},[903],{"type":47,"value":904},"\"\n",{"type":41,"tag":254,"props":906,"children":907},{"class":256,"line":613},[908],{"type":41,"tag":254,"props":909,"children":911},{"emptyLinePlaceholder":910},true,[912],{"type":47,"value":913},"\n",{"type":41,"tag":254,"props":915,"children":916},{"class":256,"line":670},[917,921,925,929,934,939,944,948,952,956],{"type":41,"tag":254,"props":918,"children":919},{"style":261},[920],{"type":47,"value":264},{"type":41,"tag":254,"props":922,"children":923},{"style":267},[924],{"type":47,"value":623},{"type":41,"tag":254,"props":926,"children":927},{"style":267},[928],{"type":47,"value":628},{"type":41,"tag":254,"props":930,"children":931},{"style":267},[932],{"type":47,"value":933}," POST",{"type":41,"tag":254,"props":935,"children":936},{"style":284},[937],{"type":47,"value":938}," \"${",{"type":41,"tag":254,"props":940,"children":941},{"style":300},[942],{"type":47,"value":943},"VSS_AGENT_BASE_URL",{"type":41,"tag":254,"props":945,"children":946},{"style":284},[947],{"type":47,"value":308},{"type":41,"tag":254,"props":949,"children":950},{"style":267},[951],{"type":47,"value":438},{"type":41,"tag":254,"props":953,"children":954},{"style":284},[955],{"type":47,"value":318},{"type":41,"tag":254,"props":957,"children":958},{"style":300},[959],{"type":47,"value":667},{"type":41,"tag":254,"props":961,"children":962},{"class":256,"line":696},[963,967,971,976,980],{"type":41,"tag":254,"props":964,"children":965},{"style":267},[966],{"type":47,"value":676},{"type":41,"tag":254,"props":968,"children":969},{"style":284},[970],{"type":47,"value":287},{"type":41,"tag":254,"props":972,"children":973},{"style":267},[974],{"type":47,"value":975},"Content-Type: application\u002Fjson",{"type":41,"tag":254,"props":977,"children":978},{"style":284},[979],{"type":47,"value":318},{"type":41,"tag":254,"props":981,"children":982},{"style":300},[983],{"type":47,"value":667},{"type":41,"tag":254,"props":985,"children":986},{"class":256,"line":721},[987,992,996,1001,1006,1010,1014],{"type":41,"tag":254,"props":988,"children":989},{"style":267},[990],{"type":47,"value":991},"  -d",{"type":41,"tag":254,"props":993,"children":994},{"style":284},[995],{"type":47,"value":520},{"type":41,"tag":254,"props":997,"children":998},{"style":267},[999],{"type":47,"value":1000},"{\"input_message\": \"Call video_understanding tool to answer the following question about \u003Csensor-id>: \u003Cuser query>\"}",{"type":41,"tag":254,"props":1002,"children":1003},{"style":284},[1004],{"type":47,"value":1005},"'",{"type":41,"tag":254,"props":1007,"children":1008},{"style":284},[1009],{"type":47,"value":510},{"type":41,"tag":254,"props":1011,"children":1012},{"style":261},[1013],{"type":47,"value":515},{"type":41,"tag":254,"props":1015,"children":1016},{"style":267},[1017],{"type":47,"value":765},{"type":41,"tag":1019,"props":1020,"children":1022},"h3",{"id":1021},"response-contract-and-extraction",[1023],{"type":47,"value":1024},"Response contract and extraction",{"type":41,"tag":50,"props":1026,"children":1027},{},[1028,1033,1035,1041],{"type":41,"tag":211,"props":1029,"children":1031},{"className":1030},[],[1032],{"type":47,"value":438},{"type":47,"value":1034}," returns a JSON object with the assistant output in ",{"type":41,"tag":211,"props":1036,"children":1038},{"className":1037},[],[1039],{"type":47,"value":1040},"value",{"type":47,"value":1042},", for example:",{"type":41,"tag":243,"props":1044,"children":1048},{"className":1045,"code":1046,"language":1047,"meta":248,"style":248},"language-json shiki shiki-themes material-theme-lighter material-theme material-theme-palenight","{\"value\":\"\u003Cagent-think>\u003Cagent-think-step ...>...\u003C\u002Fagent-think-step>\u003C\u002Fagent-think>\\n\\n\u003Cfinal answer>\\n\\n\"}\n","json",[1049],{"type":41,"tag":211,"props":1050,"children":1051},{"__ignoreMap":248},[1052],{"type":41,"tag":254,"props":1053,"children":1054},{"class":256,"line":257},[1055,1060,1064,1068,1072,1077,1081,1086,1091,1096,1100,1104],{"type":41,"tag":254,"props":1056,"children":1057},{"style":284},[1058],{"type":47,"value":1059},"{",{"type":41,"tag":254,"props":1061,"children":1062},{"style":284},[1063],{"type":47,"value":318},{"type":41,"tag":254,"props":1065,"children":1066},{"style":877},[1067],{"type":47,"value":1040},{"type":41,"tag":254,"props":1069,"children":1070},{"style":284},[1071],{"type":47,"value":318},{"type":41,"tag":254,"props":1073,"children":1074},{"style":284},[1075],{"type":47,"value":1076},":",{"type":41,"tag":254,"props":1078,"children":1079},{"style":284},[1080],{"type":47,"value":318},{"type":41,"tag":254,"props":1082,"children":1083},{"style":267},[1084],{"type":47,"value":1085},"\u003Cagent-think>\u003Cagent-think-step ...>...\u003C\u002Fagent-think-step>\u003C\u002Fagent-think>",{"type":41,"tag":254,"props":1087,"children":1088},{"style":300},[1089],{"type":47,"value":1090},"\\n\\n",{"type":41,"tag":254,"props":1092,"children":1093},{"style":267},[1094],{"type":47,"value":1095},"\u003Cfinal answer>",{"type":41,"tag":254,"props":1097,"children":1098},{"style":300},[1099],{"type":47,"value":1090},{"type":41,"tag":254,"props":1101,"children":1102},{"style":284},[1103],{"type":47,"value":318},{"type":41,"tag":254,"props":1105,"children":1106},{"style":284},[1107],{"type":47,"value":1108},"}\n",{"type":41,"tag":50,"props":1110,"children":1111},{},[1112,1114,1120,1122,1128],{"type":47,"value":1113},"There is no separate clean-answer field. The consumable answer is the text in ",{"type":41,"tag":211,"props":1115,"children":1117},{"className":1116},[],[1118],{"type":47,"value":1119},".value",{"type":47,"value":1121}," after removing any ",{"type":41,"tag":211,"props":1123,"children":1125},{"className":1124},[],[1126],{"type":47,"value":1127},"\u003Cagent-think>...\u003C\u002Fagent-think>",{"type":47,"value":1129}," block.",{"type":41,"tag":50,"props":1131,"children":1132},{},[1133],{"type":47,"value":1134},"Required handling for this skill (and any downstream caller):",{"type":41,"tag":234,"props":1136,"children":1137},{},[1138,1150,1162],{"type":41,"tag":86,"props":1139,"children":1140},{},[1141,1143,1148],{"type":47,"value":1142},"Read ",{"type":41,"tag":211,"props":1144,"children":1146},{"className":1145},[],[1147],{"type":47,"value":1119},{"type":47,"value":1149}," from the JSON response.",{"type":41,"tag":86,"props":1151,"children":1152},{},[1153,1155,1160],{"type":47,"value":1154},"Strip ",{"type":41,"tag":211,"props":1156,"children":1158},{"className":1157},[],[1159],{"type":47,"value":1127},{"type":47,"value":1161}," sections wherever they appear.",{"type":41,"tag":86,"props":1163,"children":1164},{},[1165],{"type":47,"value":1166},"Return only the remaining final-answer text to the user.",{"type":41,"tag":50,"props":1168,"children":1169},{},[1170],{"type":47,"value":1171},"Example extraction:",{"type":41,"tag":243,"props":1173,"children":1175},{"className":245,"code":1174,"language":247,"meta":248,"style":248},"curl -s -X POST \"${VSS_AGENT_BASE_URL}\u002Fgenerate\" \\\n  -H \"Content-Type: application\u002Fjson\" \\\n  -d '{\"input_message\":\"Call video_understanding tool to answer the following question about \u003Csensor-id>: \u003Cuser query>\"}' \\\n| jq -r '.value' \\\n| python3 -c 'import re,sys; t=sys.stdin.read(); t=re.sub(r\"\u003Cagent-think>.*?\u003C\u002Fagent-think>\\s*\", \"\", t, flags=re.S); print(t.strip())'\n",[1176],{"type":41,"tag":211,"props":1177,"children":1178},{"__ignoreMap":248},[1179,1222,1245,1269,1302],{"type":41,"tag":254,"props":1180,"children":1181},{"class":256,"line":257},[1182,1186,1190,1194,1198,1202,1206,1210,1214,1218],{"type":41,"tag":254,"props":1183,"children":1184},{"style":261},[1185],{"type":47,"value":264},{"type":41,"tag":254,"props":1187,"children":1188},{"style":267},[1189],{"type":47,"value":623},{"type":41,"tag":254,"props":1191,"children":1192},{"style":267},[1193],{"type":47,"value":628},{"type":41,"tag":254,"props":1195,"children":1196},{"style":267},[1197],{"type":47,"value":933},{"type":41,"tag":254,"props":1199,"children":1200},{"style":284},[1201],{"type":47,"value":938},{"type":41,"tag":254,"props":1203,"children":1204},{"style":300},[1205],{"type":47,"value":943},{"type":41,"tag":254,"props":1207,"children":1208},{"style":284},[1209],{"type":47,"value":308},{"type":41,"tag":254,"props":1211,"children":1212},{"style":267},[1213],{"type":47,"value":438},{"type":41,"tag":254,"props":1215,"children":1216},{"style":284},[1217],{"type":47,"value":318},{"type":41,"tag":254,"props":1219,"children":1220},{"style":300},[1221],{"type":47,"value":667},{"type":41,"tag":254,"props":1223,"children":1224},{"class":256,"line":604},[1225,1229,1233,1237,1241],{"type":41,"tag":254,"props":1226,"children":1227},{"style":267},[1228],{"type":47,"value":676},{"type":41,"tag":254,"props":1230,"children":1231},{"style":284},[1232],{"type":47,"value":287},{"type":41,"tag":254,"props":1234,"children":1235},{"style":267},[1236],{"type":47,"value":975},{"type":41,"tag":254,"props":1238,"children":1239},{"style":284},[1240],{"type":47,"value":318},{"type":41,"tag":254,"props":1242,"children":1243},{"style":300},[1244],{"type":47,"value":667},{"type":41,"tag":254,"props":1246,"children":1247},{"class":256,"line":613},[1248,1252,1256,1261,1265],{"type":41,"tag":254,"props":1249,"children":1250},{"style":267},[1251],{"type":47,"value":991},{"type":41,"tag":254,"props":1253,"children":1254},{"style":284},[1255],{"type":47,"value":520},{"type":41,"tag":254,"props":1257,"children":1258},{"style":267},[1259],{"type":47,"value":1260},"{\"input_message\":\"Call video_understanding tool to answer the following question about \u003Csensor-id>: \u003Cuser query>\"}",{"type":41,"tag":254,"props":1262,"children":1263},{"style":284},[1264],{"type":47,"value":1005},{"type":41,"tag":254,"props":1266,"children":1267},{"style":300},[1268],{"type":47,"value":667},{"type":41,"tag":254,"props":1270,"children":1271},{"class":256,"line":670},[1272,1277,1281,1286,1290,1294,1298],{"type":41,"tag":254,"props":1273,"children":1274},{"style":284},[1275],{"type":47,"value":1276},"|",{"type":41,"tag":254,"props":1278,"children":1279},{"style":261},[1280],{"type":47,"value":515},{"type":41,"tag":254,"props":1282,"children":1283},{"style":267},[1284],{"type":47,"value":1285}," -r",{"type":41,"tag":254,"props":1287,"children":1288},{"style":284},[1289],{"type":47,"value":520},{"type":41,"tag":254,"props":1291,"children":1292},{"style":267},[1293],{"type":47,"value":1119},{"type":41,"tag":254,"props":1295,"children":1296},{"style":284},[1297],{"type":47,"value":1005},{"type":41,"tag":254,"props":1299,"children":1300},{"style":300},[1301],{"type":47,"value":667},{"type":41,"tag":254,"props":1303,"children":1304},{"class":256,"line":696},[1305,1309,1314,1319,1323,1328],{"type":41,"tag":254,"props":1306,"children":1307},{"style":284},[1308],{"type":47,"value":1276},{"type":41,"tag":254,"props":1310,"children":1311},{"style":261},[1312],{"type":47,"value":1313}," python3",{"type":41,"tag":254,"props":1315,"children":1316},{"style":267},[1317],{"type":47,"value":1318}," -c",{"type":41,"tag":254,"props":1320,"children":1321},{"style":284},[1322],{"type":47,"value":520},{"type":41,"tag":254,"props":1324,"children":1325},{"style":267},[1326],{"type":47,"value":1327},"import re,sys; t=sys.stdin.read(); t=re.sub(r\"\u003Cagent-think>.*?\u003C\u002Fagent-think>\\s*\", \"\", t, flags=re.S); print(t.strip())",{"type":41,"tag":254,"props":1329,"children":1330},{"style":284},[1331],{"type":47,"value":530},{"type":41,"tag":71,"props":1333,"children":1334},{},[],{"type":41,"tag":75,"props":1336,"children":1338},{"id":1337},"cross-reference",[1339],{"type":47,"value":1340},"Cross-Reference",{"type":41,"tag":82,"props":1342,"children":1343},{},[1344,1365],{"type":41,"tag":86,"props":1345,"children":1346},{},[1347,1352,1354,1363],{"type":41,"tag":56,"props":1348,"children":1349},{},[1350],{"type":47,"value":1351},"vss-manage-video-io-storage",{"type":47,"value":1353}," — VST storage\u002Freplay URLs so ",{"type":41,"tag":56,"props":1355,"children":1356},{},[1357],{"type":41,"tag":211,"props":1358,"children":1360},{"className":1359},[],[1361],{"type":47,"value":1362},"VIDEO_URL",{"type":47,"value":1364}," is valid for the VLM.",{"type":41,"tag":86,"props":1366,"children":1367},{},[1368,1373,1375,1380,1382,1387,1389,1394,1396,1406,1408,1413],{"type":41,"tag":56,"props":1369,"children":1370},{},[1371],{"type":47,"value":1372},"vss-generate-video-report",{"type":47,"value":1374}," — timestamped ",{"type":41,"tag":56,"props":1376,"children":1377},{},[1378],{"type":47,"value":1379},"reports",{"type":47,"value":1381}," via ",{"type":41,"tag":56,"props":1383,"children":1384},{},[1385],{"type":47,"value":1386},"Mode A (direct VLM)",{"type":47,"value":1388}," or ",{"type":41,"tag":56,"props":1390,"children":1391},{},[1392],{"type":47,"value":1393},"Mode B (video-analytics incidents)",{"type":47,"value":1395},"; this skill is ",{"type":41,"tag":56,"props":1397,"children":1398},{},[1399,1401],{"type":47,"value":1400},"VSS-agent ",{"type":41,"tag":211,"props":1402,"children":1404},{"className":1403},[],[1405],{"type":47,"value":438},{"type":47,"value":1407}," for ad-hoc ",{"type":41,"tag":56,"props":1409,"children":1410},{},[1411],{"type":47,"value":1412},"video Q&A",{"type":47,"value":157},{"type":41,"tag":1415,"props":1416,"children":1417},"style",{},[1418],{"type":47,"value":1419},"html .light .shiki span {color: var(--shiki-light);background: var(--shiki-light-bg);font-style: var(--shiki-light-font-style);font-weight: var(--shiki-light-font-weight);text-decoration: var(--shiki-light-text-decoration);}html.light .shiki span {color: var(--shiki-light);background: var(--shiki-light-bg);font-style: var(--shiki-light-font-style);font-weight: var(--shiki-light-font-weight);text-decoration: var(--shiki-light-text-decoration);}html .default .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}html.dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}",{"items":1421,"total":1580},[1422,1440,1458,1469,1481,1495,1508,1522,1535,1546,1560,1569],{"slug":1423,"name":1423,"fn":1424,"description":1425,"org":1426,"tags":1427,"stars":1437,"repoUrl":1438,"updatedAt":1439},"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},[1428,1431,1434],{"name":1429,"slug":1430,"type":13},"Documentation","documentation",{"name":1432,"slug":1433,"type":13},"MCP","mcp",{"name":1435,"slug":1436,"type":13},"Search","search",21777,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FNemoClaw","2026-07-20T06:00:01.461044",{"slug":1441,"name":1441,"fn":1442,"description":1443,"org":1444,"tags":1445,"stars":1455,"repoUrl":1456,"updatedAt":1457},"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},[1446,1449,1452],{"name":1447,"slug":1448,"type":13},"Containers","containers",{"name":1450,"slug":1451,"type":13},"Deployment","deployment",{"name":1453,"slug":1454,"type":13},"Python","python",17049,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FMegatron-LM","2026-07-27T06:06:11.249662",{"slug":1459,"name":1459,"fn":1460,"description":1461,"org":1462,"tags":1463,"stars":1455,"repoUrl":1456,"updatedAt":1468},"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},[1464,1467],{"name":1465,"slug":1466,"type":13},"CI\u002FCD","ci-cd",{"name":1450,"slug":1451,"type":13},"2026-07-14T05:25:59.97109",{"slug":1470,"name":1470,"fn":1471,"description":1472,"org":1473,"tags":1474,"stars":1455,"repoUrl":1456,"updatedAt":1480},"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},[1475,1476,1477],{"name":1465,"slug":1466,"type":13},{"name":1450,"slug":1451,"type":13},{"name":1478,"slug":1479,"type":13},"GitHub","github","2026-07-27T06:06:12.278222",{"slug":1482,"name":1482,"fn":1483,"description":1484,"org":1485,"tags":1486,"stars":1455,"repoUrl":1456,"updatedAt":1494},"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},[1487,1490,1491],{"name":1488,"slug":1489,"type":13},"Debugging","debugging",{"name":1478,"slug":1479,"type":13},{"name":1492,"slug":1493,"type":13},"Triage","triage","2026-07-14T05:25:57.442089",{"slug":1496,"name":1496,"fn":1497,"description":1498,"org":1499,"tags":1500,"stars":1455,"repoUrl":1456,"updatedAt":1507},"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},[1501,1504],{"name":1502,"slug":1503,"type":13},"Best Practices","best-practices",{"name":1505,"slug":1506,"type":13},"Code Analysis","code-analysis","2026-07-14T05:25:56.18433",{"slug":1509,"name":1509,"fn":1510,"description":1511,"org":1512,"tags":1513,"stars":1455,"repoUrl":1456,"updatedAt":1521},"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},[1514,1517,1520],{"name":1515,"slug":1516,"type":13},"Machine Learning","machine-learning",{"name":1518,"slug":1519,"type":13},"Migration","migration",{"name":9,"slug":8,"type":13},"2026-07-17T06:07:11.777011",{"slug":1523,"name":1523,"fn":1524,"description":1525,"org":1526,"tags":1527,"stars":1455,"repoUrl":1456,"updatedAt":1534},"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},[1528,1531],{"name":1529,"slug":1530,"type":13},"QA","qa",{"name":1532,"slug":1533,"type":13},"Testing","testing","2026-07-14T05:25:53.673039",{"slug":1536,"name":1536,"fn":1537,"description":1538,"org":1539,"tags":1540,"stars":1455,"repoUrl":1456,"updatedAt":1545},"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},[1541,1542],{"name":1450,"slug":1451,"type":13},{"name":1543,"slug":1544,"type":13},"Infrastructure","infrastructure","2026-07-14T05:25:49.362534",{"slug":1547,"name":1547,"fn":1548,"description":1549,"org":1550,"tags":1551,"stars":1455,"repoUrl":1456,"updatedAt":1559},"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},[1552,1555,1556],{"name":1553,"slug":1554,"type":13},"Code Review","code-review",{"name":1478,"slug":1479,"type":13},{"name":1557,"slug":1558,"type":13},"Pull Requests","pull-requests","2026-07-14T05:26:01.226578",{"slug":1561,"name":1561,"fn":1562,"description":1563,"org":1564,"tags":1565,"stars":1455,"repoUrl":1456,"updatedAt":1568},"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},[1566,1567],{"name":1529,"slug":1530,"type":13},{"name":1532,"slug":1533,"type":13},"2026-07-14T05:25:54.928983",{"slug":1570,"name":1570,"fn":1571,"description":1572,"org":1573,"tags":1574,"stars":1455,"repoUrl":1456,"updatedAt":1579},"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},[1575,1578],{"name":1576,"slug":1577,"type":13},"Automation","automation",{"name":1465,"slug":1466,"type":13},"2026-07-30T05:29:03.275638",496,{"items":1582,"total":1676},[1583,1600,1610,1624,1634,1647,1662],{"slug":1584,"name":1584,"fn":1585,"description":1586,"org":1587,"tags":1588,"stars":20,"repoUrl":21,"updatedAt":1599},"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},[1589,1592,1595,1596],{"name":1590,"slug":1591,"type":13},"Data Analysis","data-analysis",{"name":1593,"slug":1594,"type":13},"Data Engineering","data-engineering",{"name":9,"slug":8,"type":13},{"name":1597,"slug":1598,"type":13},"Performance","performance","2026-07-14T05:28:43.176466",{"slug":1601,"name":1601,"fn":1602,"description":1603,"org":1604,"tags":1605,"stars":20,"repoUrl":21,"updatedAt":1609},"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},[1606,1607,1608],{"name":1450,"slug":1451,"type":13},{"name":1543,"slug":1544,"type":13},{"name":9,"slug":8,"type":13},"2026-07-14T05:29:06.667109",{"slug":1611,"name":1611,"fn":1612,"description":1613,"org":1614,"tags":1615,"stars":20,"repoUrl":21,"updatedAt":1623},"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},[1616,1619,1620],{"name":1617,"slug":1618,"type":13},"Agents","agents",{"name":9,"slug":8,"type":13},{"name":1621,"slug":1622,"type":13},"Research","research","2026-07-14T05:28:06.816956",{"slug":1625,"name":1625,"fn":1626,"description":1627,"org":1628,"tags":1629,"stars":20,"repoUrl":21,"updatedAt":1633},"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},[1630,1631,1632],{"name":1590,"slug":1591,"type":13},{"name":9,"slug":8,"type":13},{"name":1532,"slug":1533,"type":13},"2026-07-17T05:29:03.913266",{"slug":1635,"name":1635,"fn":1636,"description":1637,"org":1638,"tags":1639,"stars":20,"repoUrl":21,"updatedAt":1646},"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},[1640,1641,1644,1645],{"name":1576,"slug":1577,"type":13},{"name":1642,"slug":1643,"type":13},"Imaging","imaging",{"name":9,"slug":8,"type":13},{"name":18,"slug":19,"type":13},"2026-07-17T05:28:53.905004",{"slug":1648,"name":1648,"fn":1649,"description":1650,"org":1651,"tags":1652,"stars":20,"repoUrl":21,"updatedAt":1661},"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},[1653,1654,1657,1658],{"name":1450,"slug":1451,"type":13},{"name":1655,"slug":1656,"type":13},"Docker","docker",{"name":9,"slug":8,"type":13},{"name":1659,"slug":1660,"type":13},"Operations","operations","2026-07-17T05:28:56.913999",{"slug":1663,"name":1663,"fn":1664,"description":1665,"org":1666,"tags":1667,"stars":20,"repoUrl":21,"updatedAt":1675},"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},[1668,1669,1672],{"name":9,"slug":8,"type":13},{"name":1670,"slug":1671,"type":13},"Quantum Computing","quantum-computing",{"name":1673,"slug":1674,"type":13},"Simulation","simulation","2026-07-14T05:26:58.898253",305]