[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill-openai-dynamo-interconnect-check":3,"mdc--uuc62a-key":33,"related-repo-openai-dynamo-interconnect-check":1190,"related-org-openai-dynamo-interconnect-check":1313},{"slug":4,"name":4,"fn":5,"description":6,"org":7,"tags":11,"stars":22,"repoUrl":23,"updatedAt":24,"license":25,"forks":26,"topics":27,"repo":28,"sourceUrl":31,"mdContent":32},"dynamo-interconnect-check","validate Dynamo interconnect for disaggregated serving","Validate that a Dynamo deployment's NIXL\u002FUCX\u002FNCCL interconnect is ready for disaggregated serving over RDMA\u002FNVLink. Use after recipe-runner brings a deployment up (especially disagg\u002Fmulti-node) to confirm the KV transport is correct; use troubleshoot for diagnosing already-failed pods.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},"openai","OpenAI","https:\u002F\u002Fpexgzepcugksgbtrxkhf.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Forg-logos\u002Fopenai.png",[12,16,19],{"name":13,"slug":14,"type":15},"Deployment","deployment","tag",{"name":17,"slug":18,"type":15},"Engineering","engineering",{"name":20,"slug":21,"type":15},"Infrastructure","infrastructure",3992,"https:\u002F\u002Fgithub.com\u002Fopenai\u002Fplugins","2026-06-30T19:00:57.102","Apache-2.0",465,[],{"repoUrl":23,"stars":22,"forks":26,"topics":29,"description":30},[],"OpenAI Plugins","https:\u002F\u002Fgithub.com\u002Fopenai\u002Fplugins\u002Ftree\u002FHEAD\u002Fplugins\u002Fnvidia\u002Fskills\u002Fdynamo-interconnect-check","---\nname: dynamo-interconnect-check\ndescription: Validate that a Dynamo deployment's NIXL\u002FUCX\u002FNCCL interconnect is ready for disaggregated serving over RDMA\u002FNVLink. Use after recipe-runner brings a deployment up (especially disagg\u002Fmulti-node) to confirm the KV transport is correct; use troubleshoot for diagnosing already-failed pods.\nlicense: Apache-2.0\nmetadata:\n  author: Dan Gil \u003Cdagil@nvidia.com>\n  tags:\n    - dynamo\n    - nixl\n    - rdma\n    - disagg\n    - validation\n---\n\n# Dynamo Interconnect Check\n\n\u003C!--\nSPDX-FileCopyrightText: Copyright (c) 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.\nSPDX-License-Identifier: CC-BY-4.0\n-->\n\n## Purpose\n\nConfirm that the transport disaggregated serving depends on actually works. A\ndeployment can pass an endpoint smoke test while disagg is silently wrong: if\nNIXL\u002FUCX cannot reach the peer worker over RDMA or NVLink, KV transfer falls\nback to a slow or broken path. Catch that with read-only checks before trusting\na disagg deployment or its benchmark numbers.\n\nThis skill is read-only. It never mutates the cluster and never prints secrets.\n\n## Prerequisites\n\n- Python 3.10+ on the operator machine.\n- `kubectl exec` access to a worker pod in the target Dynamo deployment.\n- Read access to the recipe directory (`recipes\u002F\u003Cmodel>\u002F\u003Cframework>\u002F\u003Cmode>`).\n- For node-capability checks: tools like `ibstat`, `nvidia-smi`, `lsmod` available in the worker pod image (missing tools are reported as `skipped`, not failures).\n\n## When To Use\n\n- After `dynamo-recipe-runner` deploys a **disagg** or multi-node recipe.\n- Before reporting disagg throughput\u002Flatency, so numbers reflect the real\n  transport.\n- When agg works but disagg is slow, hangs, or returns wrong output and you\n  suspect the fabric rather than the model.\n\nFor diagnosing pods that are already crashing or unschedulable, use\n`dynamo-troubleshoot` first.\n\n## Instructions\n\n### 1. Check Transport Env Vars On The Recipe\n\n```bash\npython3 scripts\u002Fcheck_interconnect.py env recipes\u002F\u003Cmodel>\u002F\u003Cframework>\u002F\u003Cmode>\n```\n\nReports which NIXL\u002FUCX\u002FNCCL transport variables are set and flags\ndisagg-critical ones (e.g. `UCX_TLS`, `UCX_NET_DEVICES`, `NCCL_IB_HCA`) that are\nabsent. Missing here is only a warning — they may be baked into the image — so\nconfirm with the node and NIXL checks. See\n`references\u002Finterconnect-env-vars.md` for what each variable does.\n\n### 2. Check Node Capabilities\n\nLocally on a GPU node, or inside a running worker pod:\n\n```bash\npython3 scripts\u002Fcheck_interconnect.py node \\\n  --namespace \"${NAMESPACE}\" --pod \u003Cworker-pod>\n```\n\nProbes (read-only) for: InfiniBand devices and Active links, GPUDirect RDMA\n(`nvidia_peermem`), GDRCopy, and NVLink in the GPU topology. Missing tools are\nreported as `skipped`, not failures.\n\n### 3. Validate NIXL Reachability\n\n```bash\npython3 scripts\u002Fcheck_interconnect.py nixl \\\n  --namespace \"${NAMESPACE}\" --pod \u003Cworker-pod>\n```\n\nLooks for NIXL test tooling in the pod and surfaces the exact next step to run a\npairwise prefill↔decode transfer test. A full cross-pod transfer test requires\ntwo scheduled GPU pods on the fabric.\n\n## Available Scripts\n\n| Script | Purpose | Arguments |\n|---|---|---|\n| `scripts\u002Fcheck_interconnect.py env` | Inspect NIXL\u002FUCX\u002FNCCL env vars on a recipe | positional recipe path |\n| `scripts\u002Fcheck_interconnect.py node` | Probe InfiniBand, GPUDirect RDMA, GDRCopy, NVLink on a node or pod | `--namespace`, `--pod` |\n| `scripts\u002Fcheck_interconnect.py nixl` | Surface NIXL transfer-test readiness for a pod | `--namespace`, `--pod` |\n\nInvoke via the agentskills.io `run_script()` protocol:\n\n```python\nrun_script(\"scripts\u002Fcheck_interconnect.py\", args=[\"env\", \"recipes\u002Fqwen3-coder-480b\u002Fsglang\u002Fdisagg\"])\nrun_script(\"scripts\u002Fcheck_interconnect.py\", args=[\"node\", \"--namespace\", \"dynamo-demo\", \"--pod\", \"qwen-worker-0\"])\n```\n\n## Examples\n\nVerify a disagg recipe's transport env shape before deploy:\n\n```bash\npython3 scripts\u002Fcheck_interconnect.py env recipes\u002Fqwen3-coder-480b\u002Fsglang\u002Fdisagg\n```\n\nAfter deploy, validate a worker pod's fabric:\n\n```bash\npython3 scripts\u002Fcheck_interconnect.py node \\\n  --namespace dynamo-demo --pod qwen-worker-0\npython3 scripts\u002Fcheck_interconnect.py nixl \\\n  --namespace dynamo-demo --pod qwen-worker-0\n```\n\nEquivalent through the agent protocol:\n\n```python\nrun_script(\"scripts\u002Fcheck_interconnect.py\", args=[\"nixl\", \"--namespace\", \"dynamo-demo\", \"--pod\", \"qwen-worker-0\"])\n```\n\n## Output Contract\n\nEach check returns `ok` \u002F `warn` \u002F `fail` \u002F `skipped` with a one-line detail,\nplus a rolled-up verdict on disagg transport readiness. Report:\n\n- transport env vars present vs. disagg-critical ones missing\n- RDMA \u002F GPUDirect \u002F NVLink capability status\n- whether NIXL reachability was validated, and the next command if not\n- a clear statement of whether disagg can be trusted, or what to fix first\n\n## Limitations\n\n- Read-only fabric probe; does not run a full pairwise NIXL transfer (requires two scheduled GPU pods and the in-pod NIXL test tools).\n- `skipped` results for missing tools (`ibstat`, `nvidia-smi`, `lsmod`) are inconclusive, not a pass.\n- Env-var check inspects the recipe text; values injected at runtime via initContainers or operator-applied envs are not detected.\n- Single-node agg deployments do not exercise the transport — this skill is for disagg \u002F multi-node validation.\n\n## Troubleshooting\n\n| Symptom | Likely cause | Next step |\n|---|---|---|\n| `env` reports all critical vars missing | Vars baked into image or injected by operator | Run the `node` check inside the worker pod to verify actual env |\n| `node` reports no Active IB link | Fabric down or HCA not provisioned to the node | Contact cluster admin; verify `kubectl describe node` shows `nvidia.com\u002Fgpu` and IB labels |\n| `nvidia_peermem` missing | GPUDirect RDMA module not loaded | Ask cluster admin to load `nvidia-peermem`; without it, NIXL falls back to staged copies |\n| `nixl` finds no test tools | Worker image lacks NIXL test harness | Use a NIXL-enabled image or run the standalone transfer test from a debug pod |\n\n## Benchmark\n\nSee `BENCHMARK.md` for the NVCARPS-EVAL performance report (auto-generated by the NVSkills CI pipeline). To refresh, re-run `\u002Fnvskills-ci` on an upstream PR touching this skill.\n\n## References\n\n- `references\u002Finterconnect-env-vars.md` — NIXL\u002FUCX\u002FNCCL env var catalog and IB\n  capability checklist.\n- Use `scripts\u002Fcheck_interconnect.py` for all read-only checks.\n",{"data":34,"body":43},{"name":4,"description":6,"license":25,"metadata":35},{"author":36,"tags":37},"Dan Gil \u003Cdagil@nvidia.com>",[38,39,40,41,42],"dynamo","nixl","rdma","disagg","validation",{"type":44,"children":45},"root",[46,54,61,67,72,78,149,155,188,201,207,214,321,356,362,367,443,463,469,535,540,546,664,677,702,708,713,740,745,832,837,851,857,891,914,920,967,973,1125,1131,1152,1158,1184],{"type":47,"tag":48,"props":49,"children":50},"element","h1",{"id":4},[51],{"type":52,"value":53},"text","Dynamo Interconnect Check",{"type":47,"tag":55,"props":56,"children":58},"h2",{"id":57},"purpose",[59],{"type":52,"value":60},"Purpose",{"type":47,"tag":62,"props":63,"children":64},"p",{},[65],{"type":52,"value":66},"Confirm that the transport disaggregated serving depends on actually works. A\ndeployment can pass an endpoint smoke test while disagg is silently wrong: if\nNIXL\u002FUCX cannot reach the peer worker over RDMA or NVLink, KV transfer falls\nback to a slow or broken path. Catch that with read-only checks before trusting\na disagg deployment or its benchmark numbers.",{"type":47,"tag":62,"props":68,"children":69},{},[70],{"type":52,"value":71},"This skill is read-only. It never mutates the cluster and never prints secrets.",{"type":47,"tag":55,"props":73,"children":75},{"id":74},"prerequisites",[76],{"type":52,"value":77},"Prerequisites",{"type":47,"tag":79,"props":80,"children":81},"ul",{},[82,88,100,113],{"type":47,"tag":83,"props":84,"children":85},"li",{},[86],{"type":52,"value":87},"Python 3.10+ on the operator machine.",{"type":47,"tag":83,"props":89,"children":90},{},[91,98],{"type":47,"tag":92,"props":93,"children":95},"code",{"className":94},[],[96],{"type":52,"value":97},"kubectl exec",{"type":52,"value":99}," access to a worker pod in the target Dynamo deployment.",{"type":47,"tag":83,"props":101,"children":102},{},[103,105,111],{"type":52,"value":104},"Read access to the recipe directory (",{"type":47,"tag":92,"props":106,"children":108},{"className":107},[],[109],{"type":52,"value":110},"recipes\u002F\u003Cmodel>\u002F\u003Cframework>\u002F\u003Cmode>",{"type":52,"value":112},").",{"type":47,"tag":83,"props":114,"children":115},{},[116,118,124,126,132,133,139,141,147],{"type":52,"value":117},"For node-capability checks: tools like ",{"type":47,"tag":92,"props":119,"children":121},{"className":120},[],[122],{"type":52,"value":123},"ibstat",{"type":52,"value":125},", ",{"type":47,"tag":92,"props":127,"children":129},{"className":128},[],[130],{"type":52,"value":131},"nvidia-smi",{"type":52,"value":125},{"type":47,"tag":92,"props":134,"children":136},{"className":135},[],[137],{"type":52,"value":138},"lsmod",{"type":52,"value":140}," available in the worker pod image (missing tools are reported as ",{"type":47,"tag":92,"props":142,"children":144},{"className":143},[],[145],{"type":52,"value":146},"skipped",{"type":52,"value":148},", not failures).",{"type":47,"tag":55,"props":150,"children":152},{"id":151},"when-to-use",[153],{"type":52,"value":154},"When To Use",{"type":47,"tag":79,"props":156,"children":157},{},[158,178,183],{"type":47,"tag":83,"props":159,"children":160},{},[161,163,169,171,176],{"type":52,"value":162},"After ",{"type":47,"tag":92,"props":164,"children":166},{"className":165},[],[167],{"type":52,"value":168},"dynamo-recipe-runner",{"type":52,"value":170}," deploys a ",{"type":47,"tag":172,"props":173,"children":174},"strong",{},[175],{"type":52,"value":41},{"type":52,"value":177}," or multi-node recipe.",{"type":47,"tag":83,"props":179,"children":180},{},[181],{"type":52,"value":182},"Before reporting disagg throughput\u002Flatency, so numbers reflect the real\ntransport.",{"type":47,"tag":83,"props":184,"children":185},{},[186],{"type":52,"value":187},"When agg works but disagg is slow, hangs, or returns wrong output and you\nsuspect the fabric rather than the model.",{"type":47,"tag":62,"props":189,"children":190},{},[191,193,199],{"type":52,"value":192},"For diagnosing pods that are already crashing or unschedulable, use\n",{"type":47,"tag":92,"props":194,"children":196},{"className":195},[],[197],{"type":52,"value":198},"dynamo-troubleshoot",{"type":52,"value":200}," first.",{"type":47,"tag":55,"props":202,"children":204},{"id":203},"instructions",[205],{"type":52,"value":206},"Instructions",{"type":47,"tag":208,"props":209,"children":211},"h3",{"id":210},"_1-check-transport-env-vars-on-the-recipe",[212],{"type":52,"value":213},"1. Check Transport Env Vars On The Recipe",{"type":47,"tag":215,"props":216,"children":221},"pre",{"className":217,"code":218,"language":219,"meta":220,"style":220},"language-bash shiki shiki-themes material-theme-lighter material-theme material-theme-palenight","python3 scripts\u002Fcheck_interconnect.py env recipes\u002F\u003Cmodel>\u002F\u003Cframework>\u002F\u003Cmode>\n","bash","",[222],{"type":47,"tag":92,"props":223,"children":224},{"__ignoreMap":220},[225],{"type":47,"tag":226,"props":227,"children":230},"span",{"class":228,"line":229},"line",1,[231,237,243,248,253,259,264,270,275,280,284,289,294,298,302,306,311,316],{"type":47,"tag":226,"props":232,"children":234},{"style":233},"--shiki-light:#E2931D;--shiki-default:#FFCB6B;--shiki-dark:#FFCB6B",[235],{"type":52,"value":236},"python3",{"type":47,"tag":226,"props":238,"children":240},{"style":239},"--shiki-light:#91B859;--shiki-default:#C3E88D;--shiki-dark:#C3E88D",[241],{"type":52,"value":242}," scripts\u002Fcheck_interconnect.py",{"type":47,"tag":226,"props":244,"children":245},{"style":239},[246],{"type":52,"value":247}," env",{"type":47,"tag":226,"props":249,"children":250},{"style":239},[251],{"type":52,"value":252}," recipes\u002F",{"type":47,"tag":226,"props":254,"children":256},{"style":255},"--shiki-light:#39ADB5;--shiki-default:#89DDFF;--shiki-dark:#89DDFF",[257],{"type":52,"value":258},"\u003C",{"type":47,"tag":226,"props":260,"children":261},{"style":239},[262],{"type":52,"value":263},"mode",{"type":47,"tag":226,"props":265,"children":267},{"style":266},"--shiki-light:#90A4AE;--shiki-default:#EEFFFF;--shiki-dark:#BABED8",[268],{"type":52,"value":269},"l",{"type":47,"tag":226,"props":271,"children":272},{"style":255},[273],{"type":52,"value":274},">",{"type":47,"tag":226,"props":276,"children":277},{"style":239},[278],{"type":52,"value":279},"\u002F",{"type":47,"tag":226,"props":281,"children":282},{"style":255},[283],{"type":52,"value":258},{"type":47,"tag":226,"props":285,"children":286},{"style":239},[287],{"type":52,"value":288},"framewor",{"type":47,"tag":226,"props":290,"children":291},{"style":266},[292],{"type":52,"value":293},"k",{"type":47,"tag":226,"props":295,"children":296},{"style":255},[297],{"type":52,"value":274},{"type":47,"tag":226,"props":299,"children":300},{"style":239},[301],{"type":52,"value":279},{"type":47,"tag":226,"props":303,"children":304},{"style":255},[305],{"type":52,"value":258},{"type":47,"tag":226,"props":307,"children":308},{"style":239},[309],{"type":52,"value":310},"mod",{"type":47,"tag":226,"props":312,"children":313},{"style":266},[314],{"type":52,"value":315},"e",{"type":47,"tag":226,"props":317,"children":318},{"style":255},[319],{"type":52,"value":320},">\n",{"type":47,"tag":62,"props":322,"children":323},{},[324,326,332,333,339,340,346,348,354],{"type":52,"value":325},"Reports which NIXL\u002FUCX\u002FNCCL transport variables are set and flags\ndisagg-critical ones (e.g. ",{"type":47,"tag":92,"props":327,"children":329},{"className":328},[],[330],{"type":52,"value":331},"UCX_TLS",{"type":52,"value":125},{"type":47,"tag":92,"props":334,"children":336},{"className":335},[],[337],{"type":52,"value":338},"UCX_NET_DEVICES",{"type":52,"value":125},{"type":47,"tag":92,"props":341,"children":343},{"className":342},[],[344],{"type":52,"value":345},"NCCL_IB_HCA",{"type":52,"value":347},") that are\nabsent. Missing here is only a warning — they may be baked into the image — so\nconfirm with the node and NIXL checks. See\n",{"type":47,"tag":92,"props":349,"children":351},{"className":350},[],[352],{"type":52,"value":353},"references\u002Finterconnect-env-vars.md",{"type":52,"value":355}," for what each variable does.",{"type":47,"tag":208,"props":357,"children":359},{"id":358},"_2-check-node-capabilities",[360],{"type":52,"value":361},"2. Check Node Capabilities",{"type":47,"tag":62,"props":363,"children":364},{},[365],{"type":52,"value":366},"Locally on a GPU node, or inside a running worker pod:",{"type":47,"tag":215,"props":368,"children":370},{"className":217,"code":369,"language":219,"meta":220,"style":220},"python3 scripts\u002Fcheck_interconnect.py node \\\n  --namespace \"${NAMESPACE}\" --pod \u003Cworker-pod>\n",[371],{"type":47,"tag":92,"props":372,"children":373},{"__ignoreMap":220},[374,395],{"type":47,"tag":226,"props":375,"children":376},{"class":228,"line":229},[377,381,385,390],{"type":47,"tag":226,"props":378,"children":379},{"style":233},[380],{"type":52,"value":236},{"type":47,"tag":226,"props":382,"children":383},{"style":239},[384],{"type":52,"value":242},{"type":47,"tag":226,"props":386,"children":387},{"style":239},[388],{"type":52,"value":389}," node",{"type":47,"tag":226,"props":391,"children":392},{"style":266},[393],{"type":52,"value":394}," \\\n",{"type":47,"tag":226,"props":396,"children":398},{"class":228,"line":397},2,[399,404,409,414,419,424,429,434,439],{"type":47,"tag":226,"props":400,"children":401},{"style":239},[402],{"type":52,"value":403},"  --namespace",{"type":47,"tag":226,"props":405,"children":406},{"style":255},[407],{"type":52,"value":408}," \"${",{"type":47,"tag":226,"props":410,"children":411},{"style":266},[412],{"type":52,"value":413},"NAMESPACE",{"type":47,"tag":226,"props":415,"children":416},{"style":255},[417],{"type":52,"value":418},"}\"",{"type":47,"tag":226,"props":420,"children":421},{"style":239},[422],{"type":52,"value":423}," --pod",{"type":47,"tag":226,"props":425,"children":426},{"style":255},[427],{"type":52,"value":428}," \u003C",{"type":47,"tag":226,"props":430,"children":431},{"style":239},[432],{"type":52,"value":433},"worker-po",{"type":47,"tag":226,"props":435,"children":436},{"style":266},[437],{"type":52,"value":438},"d",{"type":47,"tag":226,"props":440,"children":441},{"style":255},[442],{"type":52,"value":320},{"type":47,"tag":62,"props":444,"children":445},{},[446,448,454,456,461],{"type":52,"value":447},"Probes (read-only) for: InfiniBand devices and Active links, GPUDirect RDMA\n(",{"type":47,"tag":92,"props":449,"children":451},{"className":450},[],[452],{"type":52,"value":453},"nvidia_peermem",{"type":52,"value":455},"), GDRCopy, and NVLink in the GPU topology. Missing tools are\nreported as ",{"type":47,"tag":92,"props":457,"children":459},{"className":458},[],[460],{"type":52,"value":146},{"type":52,"value":462},", not failures.",{"type":47,"tag":208,"props":464,"children":466},{"id":465},"_3-validate-nixl-reachability",[467],{"type":52,"value":468},"3. Validate NIXL Reachability",{"type":47,"tag":215,"props":470,"children":472},{"className":217,"code":471,"language":219,"meta":220,"style":220},"python3 scripts\u002Fcheck_interconnect.py nixl \\\n  --namespace \"${NAMESPACE}\" --pod \u003Cworker-pod>\n",[473],{"type":47,"tag":92,"props":474,"children":475},{"__ignoreMap":220},[476,496],{"type":47,"tag":226,"props":477,"children":478},{"class":228,"line":229},[479,483,487,492],{"type":47,"tag":226,"props":480,"children":481},{"style":233},[482],{"type":52,"value":236},{"type":47,"tag":226,"props":484,"children":485},{"style":239},[486],{"type":52,"value":242},{"type":47,"tag":226,"props":488,"children":489},{"style":239},[490],{"type":52,"value":491}," nixl",{"type":47,"tag":226,"props":493,"children":494},{"style":266},[495],{"type":52,"value":394},{"type":47,"tag":226,"props":497,"children":498},{"class":228,"line":397},[499,503,507,511,515,519,523,527,531],{"type":47,"tag":226,"props":500,"children":501},{"style":239},[502],{"type":52,"value":403},{"type":47,"tag":226,"props":504,"children":505},{"style":255},[506],{"type":52,"value":408},{"type":47,"tag":226,"props":508,"children":509},{"style":266},[510],{"type":52,"value":413},{"type":47,"tag":226,"props":512,"children":513},{"style":255},[514],{"type":52,"value":418},{"type":47,"tag":226,"props":516,"children":517},{"style":239},[518],{"type":52,"value":423},{"type":47,"tag":226,"props":520,"children":521},{"style":255},[522],{"type":52,"value":428},{"type":47,"tag":226,"props":524,"children":525},{"style":239},[526],{"type":52,"value":433},{"type":47,"tag":226,"props":528,"children":529},{"style":266},[530],{"type":52,"value":438},{"type":47,"tag":226,"props":532,"children":533},{"style":255},[534],{"type":52,"value":320},{"type":47,"tag":62,"props":536,"children":537},{},[538],{"type":52,"value":539},"Looks for NIXL test tooling in the pod and surfaces the exact next step to run a\npairwise prefill↔decode transfer test. A full cross-pod transfer test requires\ntwo scheduled GPU pods on the fabric.",{"type":47,"tag":55,"props":541,"children":543},{"id":542},"available-scripts",[544],{"type":52,"value":545},"Available Scripts",{"type":47,"tag":547,"props":548,"children":549},"table",{},[550,573],{"type":47,"tag":551,"props":552,"children":553},"thead",{},[554],{"type":47,"tag":555,"props":556,"children":557},"tr",{},[558,564,568],{"type":47,"tag":559,"props":560,"children":561},"th",{},[562],{"type":52,"value":563},"Script",{"type":47,"tag":559,"props":565,"children":566},{},[567],{"type":52,"value":60},{"type":47,"tag":559,"props":569,"children":570},{},[571],{"type":52,"value":572},"Arguments",{"type":47,"tag":574,"props":575,"children":576},"tbody",{},[577,600,633],{"type":47,"tag":555,"props":578,"children":579},{},[580,590,595],{"type":47,"tag":581,"props":582,"children":583},"td",{},[584],{"type":47,"tag":92,"props":585,"children":587},{"className":586},[],[588],{"type":52,"value":589},"scripts\u002Fcheck_interconnect.py env",{"type":47,"tag":581,"props":591,"children":592},{},[593],{"type":52,"value":594},"Inspect NIXL\u002FUCX\u002FNCCL env vars on a recipe",{"type":47,"tag":581,"props":596,"children":597},{},[598],{"type":52,"value":599},"positional recipe path",{"type":47,"tag":555,"props":601,"children":602},{},[603,612,617],{"type":47,"tag":581,"props":604,"children":605},{},[606],{"type":47,"tag":92,"props":607,"children":609},{"className":608},[],[610],{"type":52,"value":611},"scripts\u002Fcheck_interconnect.py node",{"type":47,"tag":581,"props":613,"children":614},{},[615],{"type":52,"value":616},"Probe InfiniBand, GPUDirect RDMA, GDRCopy, NVLink on a node or pod",{"type":47,"tag":581,"props":618,"children":619},{},[620,626,627],{"type":47,"tag":92,"props":621,"children":623},{"className":622},[],[624],{"type":52,"value":625},"--namespace",{"type":52,"value":125},{"type":47,"tag":92,"props":628,"children":630},{"className":629},[],[631],{"type":52,"value":632},"--pod",{"type":47,"tag":555,"props":634,"children":635},{},[636,645,650],{"type":47,"tag":581,"props":637,"children":638},{},[639],{"type":47,"tag":92,"props":640,"children":642},{"className":641},[],[643],{"type":52,"value":644},"scripts\u002Fcheck_interconnect.py nixl",{"type":47,"tag":581,"props":646,"children":647},{},[648],{"type":52,"value":649},"Surface NIXL transfer-test readiness for a pod",{"type":47,"tag":581,"props":651,"children":652},{},[653,658,659],{"type":47,"tag":92,"props":654,"children":656},{"className":655},[],[657],{"type":52,"value":625},{"type":52,"value":125},{"type":47,"tag":92,"props":660,"children":662},{"className":661},[],[663],{"type":52,"value":632},{"type":47,"tag":62,"props":665,"children":666},{},[667,669,675],{"type":52,"value":668},"Invoke via the agentskills.io ",{"type":47,"tag":92,"props":670,"children":672},{"className":671},[],[673],{"type":52,"value":674},"run_script()",{"type":52,"value":676}," protocol:",{"type":47,"tag":215,"props":678,"children":682},{"className":679,"code":680,"language":681,"meta":220,"style":220},"language-python shiki shiki-themes material-theme-lighter material-theme material-theme-palenight","run_script(\"scripts\u002Fcheck_interconnect.py\", args=[\"env\", \"recipes\u002Fqwen3-coder-480b\u002Fsglang\u002Fdisagg\"])\nrun_script(\"scripts\u002Fcheck_interconnect.py\", args=[\"node\", \"--namespace\", \"dynamo-demo\", \"--pod\", \"qwen-worker-0\"])\n","python",[683],{"type":47,"tag":92,"props":684,"children":685},{"__ignoreMap":220},[686,694],{"type":47,"tag":226,"props":687,"children":688},{"class":228,"line":229},[689],{"type":47,"tag":226,"props":690,"children":691},{},[692],{"type":52,"value":693},"run_script(\"scripts\u002Fcheck_interconnect.py\", args=[\"env\", \"recipes\u002Fqwen3-coder-480b\u002Fsglang\u002Fdisagg\"])\n",{"type":47,"tag":226,"props":695,"children":696},{"class":228,"line":397},[697],{"type":47,"tag":226,"props":698,"children":699},{},[700],{"type":52,"value":701},"run_script(\"scripts\u002Fcheck_interconnect.py\", args=[\"node\", \"--namespace\", \"dynamo-demo\", \"--pod\", \"qwen-worker-0\"])\n",{"type":47,"tag":55,"props":703,"children":705},{"id":704},"examples",[706],{"type":52,"value":707},"Examples",{"type":47,"tag":62,"props":709,"children":710},{},[711],{"type":52,"value":712},"Verify a disagg recipe's transport env shape before deploy:",{"type":47,"tag":215,"props":714,"children":716},{"className":217,"code":715,"language":219,"meta":220,"style":220},"python3 scripts\u002Fcheck_interconnect.py env recipes\u002Fqwen3-coder-480b\u002Fsglang\u002Fdisagg\n",[717],{"type":47,"tag":92,"props":718,"children":719},{"__ignoreMap":220},[720],{"type":47,"tag":226,"props":721,"children":722},{"class":228,"line":229},[723,727,731,735],{"type":47,"tag":226,"props":724,"children":725},{"style":233},[726],{"type":52,"value":236},{"type":47,"tag":226,"props":728,"children":729},{"style":239},[730],{"type":52,"value":242},{"type":47,"tag":226,"props":732,"children":733},{"style":239},[734],{"type":52,"value":247},{"type":47,"tag":226,"props":736,"children":737},{"style":239},[738],{"type":52,"value":739}," recipes\u002Fqwen3-coder-480b\u002Fsglang\u002Fdisagg\n",{"type":47,"tag":62,"props":741,"children":742},{},[743],{"type":52,"value":744},"After deploy, validate a worker pod's fabric:",{"type":47,"tag":215,"props":746,"children":748},{"className":217,"code":747,"language":219,"meta":220,"style":220},"python3 scripts\u002Fcheck_interconnect.py node \\\n  --namespace dynamo-demo --pod qwen-worker-0\npython3 scripts\u002Fcheck_interconnect.py nixl \\\n  --namespace dynamo-demo --pod qwen-worker-0\n",[749],{"type":47,"tag":92,"props":750,"children":751},{"__ignoreMap":220},[752,771,792,812],{"type":47,"tag":226,"props":753,"children":754},{"class":228,"line":229},[755,759,763,767],{"type":47,"tag":226,"props":756,"children":757},{"style":233},[758],{"type":52,"value":236},{"type":47,"tag":226,"props":760,"children":761},{"style":239},[762],{"type":52,"value":242},{"type":47,"tag":226,"props":764,"children":765},{"style":239},[766],{"type":52,"value":389},{"type":47,"tag":226,"props":768,"children":769},{"style":266},[770],{"type":52,"value":394},{"type":47,"tag":226,"props":772,"children":773},{"class":228,"line":397},[774,778,783,787],{"type":47,"tag":226,"props":775,"children":776},{"style":239},[777],{"type":52,"value":403},{"type":47,"tag":226,"props":779,"children":780},{"style":239},[781],{"type":52,"value":782}," dynamo-demo",{"type":47,"tag":226,"props":784,"children":785},{"style":239},[786],{"type":52,"value":423},{"type":47,"tag":226,"props":788,"children":789},{"style":239},[790],{"type":52,"value":791}," qwen-worker-0\n",{"type":47,"tag":226,"props":793,"children":795},{"class":228,"line":794},3,[796,800,804,808],{"type":47,"tag":226,"props":797,"children":798},{"style":233},[799],{"type":52,"value":236},{"type":47,"tag":226,"props":801,"children":802},{"style":239},[803],{"type":52,"value":242},{"type":47,"tag":226,"props":805,"children":806},{"style":239},[807],{"type":52,"value":491},{"type":47,"tag":226,"props":809,"children":810},{"style":266},[811],{"type":52,"value":394},{"type":47,"tag":226,"props":813,"children":815},{"class":228,"line":814},4,[816,820,824,828],{"type":47,"tag":226,"props":817,"children":818},{"style":239},[819],{"type":52,"value":403},{"type":47,"tag":226,"props":821,"children":822},{"style":239},[823],{"type":52,"value":782},{"type":47,"tag":226,"props":825,"children":826},{"style":239},[827],{"type":52,"value":423},{"type":47,"tag":226,"props":829,"children":830},{"style":239},[831],{"type":52,"value":791},{"type":47,"tag":62,"props":833,"children":834},{},[835],{"type":52,"value":836},"Equivalent through the agent protocol:",{"type":47,"tag":215,"props":838,"children":840},{"className":679,"code":839,"language":681,"meta":220,"style":220},"run_script(\"scripts\u002Fcheck_interconnect.py\", args=[\"nixl\", \"--namespace\", \"dynamo-demo\", \"--pod\", \"qwen-worker-0\"])\n",[841],{"type":47,"tag":92,"props":842,"children":843},{"__ignoreMap":220},[844],{"type":47,"tag":226,"props":845,"children":846},{"class":228,"line":229},[847],{"type":47,"tag":226,"props":848,"children":849},{},[850],{"type":52,"value":839},{"type":47,"tag":55,"props":852,"children":854},{"id":853},"output-contract",[855],{"type":52,"value":856},"Output Contract",{"type":47,"tag":62,"props":858,"children":859},{},[860,862,868,870,876,877,883,884,889],{"type":52,"value":861},"Each check returns ",{"type":47,"tag":92,"props":863,"children":865},{"className":864},[],[866],{"type":52,"value":867},"ok",{"type":52,"value":869}," \u002F ",{"type":47,"tag":92,"props":871,"children":873},{"className":872},[],[874],{"type":52,"value":875},"warn",{"type":52,"value":869},{"type":47,"tag":92,"props":878,"children":880},{"className":879},[],[881],{"type":52,"value":882},"fail",{"type":52,"value":869},{"type":47,"tag":92,"props":885,"children":887},{"className":886},[],[888],{"type":52,"value":146},{"type":52,"value":890}," with a one-line detail,\nplus a rolled-up verdict on disagg transport readiness. Report:",{"type":47,"tag":79,"props":892,"children":893},{},[894,899,904,909],{"type":47,"tag":83,"props":895,"children":896},{},[897],{"type":52,"value":898},"transport env vars present vs. disagg-critical ones missing",{"type":47,"tag":83,"props":900,"children":901},{},[902],{"type":52,"value":903},"RDMA \u002F GPUDirect \u002F NVLink capability status",{"type":47,"tag":83,"props":905,"children":906},{},[907],{"type":52,"value":908},"whether NIXL reachability was validated, and the next command if not",{"type":47,"tag":83,"props":910,"children":911},{},[912],{"type":52,"value":913},"a clear statement of whether disagg can be trusted, or what to fix first",{"type":47,"tag":55,"props":915,"children":917},{"id":916},"limitations",[918],{"type":52,"value":919},"Limitations",{"type":47,"tag":79,"props":921,"children":922},{},[923,928,957,962],{"type":47,"tag":83,"props":924,"children":925},{},[926],{"type":52,"value":927},"Read-only fabric probe; does not run a full pairwise NIXL transfer (requires two scheduled GPU pods and the in-pod NIXL test tools).",{"type":47,"tag":83,"props":929,"children":930},{},[931,936,938,943,944,949,950,955],{"type":47,"tag":92,"props":932,"children":934},{"className":933},[],[935],{"type":52,"value":146},{"type":52,"value":937}," results for missing tools (",{"type":47,"tag":92,"props":939,"children":941},{"className":940},[],[942],{"type":52,"value":123},{"type":52,"value":125},{"type":47,"tag":92,"props":945,"children":947},{"className":946},[],[948],{"type":52,"value":131},{"type":52,"value":125},{"type":47,"tag":92,"props":951,"children":953},{"className":952},[],[954],{"type":52,"value":138},{"type":52,"value":956},") are inconclusive, not a pass.",{"type":47,"tag":83,"props":958,"children":959},{},[960],{"type":52,"value":961},"Env-var check inspects the recipe text; values injected at runtime via initContainers or operator-applied envs are not detected.",{"type":47,"tag":83,"props":963,"children":964},{},[965],{"type":52,"value":966},"Single-node agg deployments do not exercise the transport — this skill is for disagg \u002F multi-node validation.",{"type":47,"tag":55,"props":968,"children":970},{"id":969},"troubleshooting",[971],{"type":52,"value":972},"Troubleshooting",{"type":47,"tag":547,"props":974,"children":975},{},[976,997],{"type":47,"tag":551,"props":977,"children":978},{},[979],{"type":47,"tag":555,"props":980,"children":981},{},[982,987,992],{"type":47,"tag":559,"props":983,"children":984},{},[985],{"type":52,"value":986},"Symptom",{"type":47,"tag":559,"props":988,"children":989},{},[990],{"type":52,"value":991},"Likely cause",{"type":47,"tag":559,"props":993,"children":994},{},[995],{"type":52,"value":996},"Next step",{"type":47,"tag":574,"props":998,"children":999},{},[1000,1032,1071,1102],{"type":47,"tag":555,"props":1001,"children":1002},{},[1003,1014,1019],{"type":47,"tag":581,"props":1004,"children":1005},{},[1006,1012],{"type":47,"tag":92,"props":1007,"children":1009},{"className":1008},[],[1010],{"type":52,"value":1011},"env",{"type":52,"value":1013}," reports all critical vars missing",{"type":47,"tag":581,"props":1015,"children":1016},{},[1017],{"type":52,"value":1018},"Vars baked into image or injected by operator",{"type":47,"tag":581,"props":1020,"children":1021},{},[1022,1024,1030],{"type":52,"value":1023},"Run the ",{"type":47,"tag":92,"props":1025,"children":1027},{"className":1026},[],[1028],{"type":52,"value":1029},"node",{"type":52,"value":1031}," check inside the worker pod to verify actual env",{"type":47,"tag":555,"props":1033,"children":1034},{},[1035,1045,1050],{"type":47,"tag":581,"props":1036,"children":1037},{},[1038,1043],{"type":47,"tag":92,"props":1039,"children":1041},{"className":1040},[],[1042],{"type":52,"value":1029},{"type":52,"value":1044}," reports no Active IB link",{"type":47,"tag":581,"props":1046,"children":1047},{},[1048],{"type":52,"value":1049},"Fabric down or HCA not provisioned to the node",{"type":47,"tag":581,"props":1051,"children":1052},{},[1053,1055,1061,1063,1069],{"type":52,"value":1054},"Contact cluster admin; verify ",{"type":47,"tag":92,"props":1056,"children":1058},{"className":1057},[],[1059],{"type":52,"value":1060},"kubectl describe node",{"type":52,"value":1062}," shows ",{"type":47,"tag":92,"props":1064,"children":1066},{"className":1065},[],[1067],{"type":52,"value":1068},"nvidia.com\u002Fgpu",{"type":52,"value":1070}," and IB labels",{"type":47,"tag":555,"props":1072,"children":1073},{},[1074,1084,1089],{"type":47,"tag":581,"props":1075,"children":1076},{},[1077,1082],{"type":47,"tag":92,"props":1078,"children":1080},{"className":1079},[],[1081],{"type":52,"value":453},{"type":52,"value":1083}," missing",{"type":47,"tag":581,"props":1085,"children":1086},{},[1087],{"type":52,"value":1088},"GPUDirect RDMA module not loaded",{"type":47,"tag":581,"props":1090,"children":1091},{},[1092,1094,1100],{"type":52,"value":1093},"Ask cluster admin to load ",{"type":47,"tag":92,"props":1095,"children":1097},{"className":1096},[],[1098],{"type":52,"value":1099},"nvidia-peermem",{"type":52,"value":1101},"; without it, NIXL falls back to staged copies",{"type":47,"tag":555,"props":1103,"children":1104},{},[1105,1115,1120],{"type":47,"tag":581,"props":1106,"children":1107},{},[1108,1113],{"type":47,"tag":92,"props":1109,"children":1111},{"className":1110},[],[1112],{"type":52,"value":39},{"type":52,"value":1114}," finds no test tools",{"type":47,"tag":581,"props":1116,"children":1117},{},[1118],{"type":52,"value":1119},"Worker image lacks NIXL test harness",{"type":47,"tag":581,"props":1121,"children":1122},{},[1123],{"type":52,"value":1124},"Use a NIXL-enabled image or run the standalone transfer test from a debug pod",{"type":47,"tag":55,"props":1126,"children":1128},{"id":1127},"benchmark",[1129],{"type":52,"value":1130},"Benchmark",{"type":47,"tag":62,"props":1132,"children":1133},{},[1134,1136,1142,1144,1150],{"type":52,"value":1135},"See ",{"type":47,"tag":92,"props":1137,"children":1139},{"className":1138},[],[1140],{"type":52,"value":1141},"BENCHMARK.md",{"type":52,"value":1143}," for the NVCARPS-EVAL performance report (auto-generated by the NVSkills CI pipeline). To refresh, re-run ",{"type":47,"tag":92,"props":1145,"children":1147},{"className":1146},[],[1148],{"type":52,"value":1149},"\u002Fnvskills-ci",{"type":52,"value":1151}," on an upstream PR touching this skill.",{"type":47,"tag":55,"props":1153,"children":1155},{"id":1154},"references",[1156],{"type":52,"value":1157},"References",{"type":47,"tag":79,"props":1159,"children":1160},{},[1161,1171],{"type":47,"tag":83,"props":1162,"children":1163},{},[1164,1169],{"type":47,"tag":92,"props":1165,"children":1167},{"className":1166},[],[1168],{"type":52,"value":353},{"type":52,"value":1170}," — NIXL\u002FUCX\u002FNCCL env var catalog and IB\ncapability checklist.",{"type":47,"tag":83,"props":1172,"children":1173},{},[1174,1176,1182],{"type":52,"value":1175},"Use ",{"type":47,"tag":92,"props":1177,"children":1179},{"className":1178},[],[1180],{"type":52,"value":1181},"scripts\u002Fcheck_interconnect.py",{"type":52,"value":1183}," for all read-only checks.",{"type":47,"tag":1185,"props":1186,"children":1187},"style",{},[1188],{"type":52,"value":1189},"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":1191,"total":1312},[1192,1210,1226,1238,1258,1280,1300],{"slug":1193,"name":1193,"fn":1194,"description":1195,"org":1196,"tags":1197,"stars":22,"repoUrl":23,"updatedAt":24},"accessibility-and-inclusive-visualization","make data visualizations accessible","Make data visualizations accessible and inclusive. Use when the user needs chart or diagram accessibility guidance, text alternatives for complex visuals, color and contrast review, keyboard support, reduced-motion behavior for animation or parallax, or an accessibility QA workflow for exported figures, UML-like diagrams, and dashboards.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1198,1201,1204,1207],{"name":1199,"slug":1200,"type":15},"Accessibility","accessibility",{"name":1202,"slug":1203,"type":15},"Charts","charts",{"name":1205,"slug":1206,"type":15},"Data Visualization","data-visualization",{"name":1208,"slug":1209,"type":15},"Design","design",{"slug":1211,"name":1211,"fn":1212,"description":1213,"org":1214,"tags":1215,"stars":22,"repoUrl":23,"updatedAt":1225},"agent-browser","automate browser interactions for agents","Browser automation CLI for AI agents. Use when the user needs to interact with websites, verify dev server output, test web apps, navigate pages, fill forms, click buttons, take screenshots, extract data, or automate any browser task. Also triggers when a dev server starts so you can verify it visually.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1216,1219,1222],{"name":1217,"slug":1218,"type":15},"Agents","agents",{"name":1220,"slug":1221,"type":15},"Browser Automation","browser-automation",{"name":1223,"slug":1224,"type":15},"Testing","testing","2026-04-06T18:41:03.44016",{"slug":1227,"name":1227,"fn":1228,"description":1229,"org":1230,"tags":1231,"stars":22,"repoUrl":23,"updatedAt":1237},"agent-browser-verify","verify dev server output with automated browser","Automated browser verification for dev servers. Triggers when a dev server starts to run a visual gut-check with agent-browser — verifies the page loads, checks for console errors, validates key UI elements, and reports pass\u002Ffail before continuing.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1232,1233,1236],{"name":1220,"slug":1221,"type":15},{"name":1234,"slug":1235,"type":15},"Local Development","local-development",{"name":1223,"slug":1224,"type":15},"2026-04-06T18:41:17.526867",{"slug":1239,"name":1239,"fn":1240,"description":1241,"org":1242,"tags":1243,"stars":22,"repoUrl":23,"updatedAt":1257},"agents-sdk","build AI agents on Cloudflare Workers","Build AI agents on Cloudflare Workers using the Agents SDK. Load when creating stateful agents, durable workflows, real-time WebSocket apps, scheduled tasks, MCP servers, or chat applications. Covers Agent class, state management, callable RPC, Workflows integration, and React hooks. Biases towards retrieval from Cloudflare docs over pre-trained knowledge.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1244,1245,1248,1251,1254],{"name":1217,"slug":1218,"type":15},{"name":1246,"slug":1247,"type":15},"Cloudflare Workers","cloudflare-workers",{"name":1249,"slug":1250,"type":15},"SDK","sdk",{"name":1252,"slug":1253,"type":15},"Serverless","serverless",{"name":1255,"slug":1256,"type":15},"WebSockets","websockets","2026-04-06T18:39:51.717063",{"slug":1259,"name":1259,"fn":1260,"description":1261,"org":1262,"tags":1263,"stars":22,"repoUrl":23,"updatedAt":1279},"ai-elements","build chat UIs with AI Elements","AI Elements component library guidance — pre-built React components for AI interfaces built on shadcn\u002Fui. Use when building chat UIs, message displays, tool call rendering, streaming responses, reasoning panels, or any AI-native interface with the AI SDK.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1264,1267,1270,1273,1276],{"name":1265,"slug":1266,"type":15},"Frontend","frontend",{"name":1268,"slug":1269,"type":15},"React","react",{"name":1271,"slug":1272,"type":15},"shadcn\u002Fui","shadcn-ui",{"name":1274,"slug":1275,"type":15},"UI Components","ui-components",{"name":1277,"slug":1278,"type":15},"Vercel","vercel","2026-04-06T18:40:59.619419",{"slug":1281,"name":1281,"fn":1282,"description":1283,"org":1284,"tags":1285,"stars":22,"repoUrl":23,"updatedAt":1299},"ai-gateway","configure Vercel AI Gateway","Vercel AI Gateway expert guidance. Use when configuring model routing, provider failover, cost tracking, or managing multiple AI providers through a unified API.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1286,1289,1292,1295,1298],{"name":1287,"slug":1288,"type":15},"AI Infrastructure","ai-infrastructure",{"name":1290,"slug":1291,"type":15},"Cost Optimization","cost-optimization",{"name":1293,"slug":1294,"type":15},"LLM","llm",{"name":1296,"slug":1297,"type":15},"Performance","performance",{"name":1277,"slug":1278,"type":15},"2026-04-06T18:40:44.377464",{"slug":1301,"name":1301,"fn":1302,"description":1303,"org":1304,"tags":1305,"stars":22,"repoUrl":23,"updatedAt":1311},"ai-generation-persistence","implement persistence patterns for AI generations","AI generation persistence patterns — unique IDs, addressable URLs, database storage, and cost tracking for every LLM generation",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1306,1307,1310],{"name":1290,"slug":1291,"type":15},{"name":1308,"slug":1309,"type":15},"Database","database",{"name":1293,"slug":1294,"type":15},"2026-04-06T18:41:08.513425",600,{"items":1314,"total":1509},[1315,1336,1359,1376,1392,1407,1426,1438,1452,1466,1478,1493],{"slug":1316,"name":1316,"fn":1317,"description":1318,"org":1319,"tags":1320,"stars":1333,"repoUrl":1334,"updatedAt":1335},"prior-auth-packet-builder","build healthcare prior authorization packets","Build a concise prior authorization packet from local case files and payer policy docs.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1321,1324,1327,1330],{"name":1322,"slug":1323,"type":15},"Documents","documents",{"name":1325,"slug":1326,"type":15},"Healthcare","healthcare",{"name":1328,"slug":1329,"type":15},"Insurance","insurance",{"name":1331,"slug":1332,"type":15},"Regulatory Compliance","regulatory-compliance",28169,"https:\u002F\u002Fgithub.com\u002Fopenai\u002Fopenai-agents-python","2026-04-16T05:11:39.180399",{"slug":1337,"name":1337,"fn":1338,"description":1339,"org":1340,"tags":1341,"stars":1356,"repoUrl":1357,"updatedAt":1358},"aspnet-core","build ASP.NET Core web applications","Build, review, refactor, or architect ASP.NET Core web applications using current official guidance for .NET web development. Use when working on Blazor Web Apps, Razor Pages, MVC, Minimal APIs, controller-based Web APIs, SignalR, gRPC, middleware, dependency injection, configuration, authentication, authorization, testing, performance, deployment, or ASP.NET Core upgrades.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1342,1345,1347,1350,1353],{"name":1343,"slug":1344,"type":15},".NET","dotnet",{"name":1346,"slug":1337,"type":15},"ASP.NET Core",{"name":1348,"slug":1349,"type":15},"Blazor","blazor",{"name":1351,"slug":1352,"type":15},"C#","csharp",{"name":1354,"slug":1355,"type":15},"Web Development","web-development",23787,"https:\u002F\u002Fgithub.com\u002Fopenai\u002Fskills","2026-04-12T05:07:02.819491",{"slug":1360,"name":1360,"fn":1361,"description":1362,"org":1363,"tags":1364,"stars":1356,"repoUrl":1357,"updatedAt":1375},"chatgpt-apps","build ChatGPT Apps SDK applications","Build, scaffold, refactor, and troubleshoot ChatGPT Apps SDK applications that combine an MCP server and widget UI. Use when Codex needs to design tools, register UI resources, wire the MCP Apps bridge or ChatGPT compatibility APIs, apply Apps SDK metadata or CSP or domain settings, or produce a docs-aligned project scaffold. Prefer a docs-first workflow by invoking the openai-docs skill or OpenAI developer docs MCP tools before generating code.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1365,1368,1371,1374],{"name":1366,"slug":1367,"type":15},"Apps SDK","apps-sdk",{"name":1369,"slug":1370,"type":15},"ChatGPT","chatgpt",{"name":1372,"slug":1373,"type":15},"MCP","mcp",{"name":9,"slug":8,"type":15},"2026-04-12T05:07:05.468097",{"slug":1377,"name":1377,"fn":1378,"description":1379,"org":1380,"tags":1381,"stars":1356,"repoUrl":1357,"updatedAt":1391},"cli-creator","build CLIs from API docs","Build a composable CLI for Codex from API docs, an OpenAPI spec, existing curl examples, an SDK, a web app, an admin tool, or a local script. Use when the user wants Codex to create a command-line tool that can run from any repo, expose composable read\u002Fwrite commands, return stable JSON, manage auth, and pair with a companion skill.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1382,1385,1388],{"name":1383,"slug":1384,"type":15},"API Development","api-development",{"name":1386,"slug":1387,"type":15},"CLI","cli",{"name":1389,"slug":1390,"type":15},"Codex","codex","2026-04-12T05:07:04.132762",{"slug":1393,"name":1393,"fn":1394,"description":1395,"org":1396,"tags":1397,"stars":1356,"repoUrl":1357,"updatedAt":1406},"cloudflare-deploy","deploy projects to Cloudflare","Deploy applications and infrastructure to Cloudflare using Workers, Pages, and related platform services. Use when the user asks to deploy, host, publish, or set up a project on Cloudflare.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1398,1401,1404,1405],{"name":1399,"slug":1400,"type":15},"Cloudflare","cloudflare",{"name":1402,"slug":1403,"type":15},"Cloudflare Pages","cloudflare-pages",{"name":1246,"slug":1247,"type":15},{"name":13,"slug":14,"type":15},"2026-04-12T05:07:14.275118",{"slug":1408,"name":1408,"fn":1409,"description":1410,"org":1411,"tags":1412,"stars":1356,"repoUrl":1357,"updatedAt":1425},"define-goal","define and set measurable project goals","Help the user define a concrete, measurable goal before starting work, especially when they ask to use the goal tool, create a goal, set an objective, clarify success criteria, or turn a fuzzy intention into a quantitative outcome. Use this skill for goal creation and goal refinement only; it does not manage durable snapshots, decision logs, or long-running execution artifacts.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1413,1416,1419,1422],{"name":1414,"slug":1415,"type":15},"Productivity","productivity",{"name":1417,"slug":1418,"type":15},"Project Management","project-management",{"name":1420,"slug":1421,"type":15},"Strategy","strategy",{"name":1423,"slug":1424,"type":15},"Task Management","task-management","2026-05-23T06:17:16.870838",{"slug":1427,"name":1427,"fn":1428,"description":1429,"org":1430,"tags":1431,"stars":1356,"repoUrl":1357,"updatedAt":1437},"figma","translate Figma designs into code","Use the Figma MCP server to fetch design context, screenshots, variables, and assets from Figma, and to translate Figma nodes into production code. Trigger when a task involves Figma URLs, node IDs, design-to-code implementation, or Figma MCP setup and troubleshooting.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1432,1433,1435,1436],{"name":1208,"slug":1209,"type":15},{"name":1434,"slug":1427,"type":15},"Figma",{"name":1265,"slug":1266,"type":15},{"name":1372,"slug":1373,"type":15},"2026-04-12T05:06:47.939943",{"slug":1439,"name":1439,"fn":1440,"description":1441,"org":1442,"tags":1443,"stars":1356,"repoUrl":1357,"updatedAt":1451},"figma-code-connect-components","connect Figma designs to code components","Connects Figma design components to code components using Code Connect mapping tools. Use when user says \"code connect\", \"connect this component to code\", \"map this component\", \"link component to code\", \"create code connect mapping\", or wants to establish mappings between Figma designs and code implementations. For canvas writes via `use_figma`, use `figma-use`.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1444,1445,1448,1449,1450],{"name":1208,"slug":1209,"type":15},{"name":1446,"slug":1447,"type":15},"Design System","design-system",{"name":1434,"slug":1427,"type":15},{"name":1265,"slug":1266,"type":15},{"name":1274,"slug":1275,"type":15},"2026-05-10T05:59:52.971881",{"slug":1453,"name":1453,"fn":1454,"description":1455,"org":1456,"tags":1457,"stars":1356,"repoUrl":1357,"updatedAt":1465},"figma-create-design-system-rules","generate design system rules from Figma","Generates custom design system rules for the user's codebase. Use when user says \"create design system rules\", \"generate rules for my project\", \"set up design rules\", \"customize design system guidelines\", or wants to establish project-specific conventions for Figma-to-code workflows. Requires Figma MCP server connection.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1458,1459,1460,1463,1464],{"name":1208,"slug":1209,"type":15},{"name":1446,"slug":1447,"type":15},{"name":1461,"slug":1462,"type":15},"Documentation","documentation",{"name":1434,"slug":1427,"type":15},{"name":1265,"slug":1266,"type":15},"2026-05-16T06:07:47.821474",{"slug":1467,"name":1467,"fn":1468,"description":1469,"org":1470,"tags":1471,"stars":1356,"repoUrl":1357,"updatedAt":1477},"figma-implement-design","translate Figma designs into application code","Translates Figma designs into production-ready application code with 1:1 visual fidelity. Use when implementing UI code from Figma files, when user mentions \"implement design\", \"generate code\", \"implement component\", provides Figma URLs, or asks to build components matching Figma specs. For Figma canvas writes via `use_figma`, use `figma-use`.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1472,1473,1474,1475,1476],{"name":1208,"slug":1209,"type":15},{"name":1434,"slug":1427,"type":15},{"name":1265,"slug":1266,"type":15},{"name":1274,"slug":1275,"type":15},{"name":1354,"slug":1355,"type":15},"2026-05-16T06:07:40.583615",{"slug":1479,"name":1479,"fn":1480,"description":1481,"org":1482,"tags":1483,"stars":1356,"repoUrl":1357,"updatedAt":1492},"hatch-pet","create animated pets for Codex","Create, repair, validate, visually QA, and package Codex-compatible animated pets and pet spritesheets from character art, generated images, company or prospect brand cues, or visual references. Use when a user wants a lightweight-worker Codex pet workflow, a non-pixel custom pet style, a prospect or company mascot pet, or a full 8x9 animated pet atlas with transparent unused cells, QA contact sheets, and pet.json packaging. This skill composes the installed $imagegen system skill for visual generation and uses bundled scripts for deterministic spritesheet assembly.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1484,1487,1488,1491],{"name":1485,"slug":1486,"type":15},"Animation","animation",{"name":1389,"slug":1390,"type":15},{"name":1489,"slug":1490,"type":15},"Creative","creative",{"name":1208,"slug":1209,"type":15},"2026-05-02T05:31:48.48485",{"slug":1494,"name":1494,"fn":1495,"description":1496,"org":1497,"tags":1498,"stars":1356,"repoUrl":1357,"updatedAt":1508},"imagegen","generate and edit raster images","Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, and the output should be a bitmap asset rather than repo-native code or vector. Do not use when the task is better handled by editing existing SVG\u002Fvector\u002Fcode-native assets, extending an established icon or logo system, or building the visual directly in HTML\u002FCSS\u002Fcanvas.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1499,1500,1501,1504,1507],{"name":1489,"slug":1490,"type":15},{"name":1208,"slug":1209,"type":15},{"name":1502,"slug":1503,"type":15},"Image Generation","image-generation",{"name":1505,"slug":1506,"type":15},"Images","images",{"name":9,"slug":8,"type":15},"2026-05-15T06:23:24.312127",675]