[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7CSJAG8377NPHC9NBTPK02P":3,"skill-same:01M7CSJAG8377NPHC9NBTPK02P":73},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":30,"commitSha":31,"skillDirectory":32,"forksCount":33,"repositoryHtmlUrl":34,"repositoryHomepageUrl":35,"ownerHtmlUrl":36,"ownerBlogUrl":6,"ownerDescription":37,"updatedAt":38,"install":39,"resources":52},"01M7CSJAG8377NPHC9NBTPK02P","nemo-automodel-launcher-config",null,"Configure NeMo AutoModel job launches for interactive runs, Slurm clusters, and SkyPilot cloud execution.","YYC-Cube\u002FYYC3-Knowledge-Base","YYC-Cube","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F235006543?v=4","01M79J2PH4SWD2X0825X6PN748","yyc3","01M79J2PHVK3F6ZBYT8K2SCZCW",1,"Apache-2.0","https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fnemo-automodel-launcher-config\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: nemo-automodel-launcher-config\ndescription: Configure NeMo AutoModel job launches for interactive runs, Slurm clusters, and SkyPilot cloud execution.\nwhen_to_use: Configuring Slurm or SkyPilot job submission, setting up multi-node launch scripts, debugging job submission failures, or switching between interactive and cluster launch modes.\nlicense: Apache-2.0\nmetadata:\n  author: NVIDIA\n  tags:\n    - nemo-automodel\n    - launcher-config\n---\n\n# Launcher Configuration\n\nNeMo AutoModel supports three launch methods: interactive (torchrun), Slurm (HPC clusters), and SkyPilot (cloud-agnostic).\n\n## Instructions\n\nFor launcher questions, answer directly from this skill without inspecting the\nrepository unless the user asks you to edit files. Keep the answer focused on\nthe relevant launch YAML, required fields, and the expected runtime behavior.\n\nUse these compact answer patterns for common questions:\n\n- Slurm multi-node: show a `slurm:` YAML block with `job_name`, `nodes`,\n  `ntasks_per_node`, `time`, `account` or `partition`, `container_image`,\n  `hf_home`, optional `extra_mounts`, `env_vars`, and `master_port`; explain\n  that the launcher derives `WORLD_SIZE = nodes * ntasks_per_node` and sets\n  `MASTER_ADDR` and `MASTER_PORT`.\n- SkyPilot spot: show a `skypilot:` YAML block with `cloud`, `accelerators`,\n  `num_nodes`, `use_spot: true`, `disk_size`, `region`, `setup`, and\n  `env_vars`; warn that spot instances can be preempted, set a short\n  `step_scheduler.checkpoint_interval`, and resume with `restore_from.path`.\n- Nsight Systems on Slurm: show `slurm.nsys_enabled: true` alongside normal\n  Slurm fields, say the launcher wraps the training command with\n  `nsys profile`, and state that it produces a `.nsys-rep` report file.\n  Treat profiling as diagnostic-only: use short profiling runs and disable it\n  for normal production training because it adds overhead and large artifacts.\n\nFor Slurm answers, start with this minimal template and then adjust only the\nfields the user asked about:\n\n```yaml\nslurm:\n  job_name: llm_finetune\n  nodes: 2\n  ntasks_per_node: 8\n  time: \"04:00:00\"\n  account: my_account\n  partition: batch\n  container_image: nvcr.io\u002Fnvidia\u002Fnemo:dev\n  hf_home: ~\u002F.cache\u002Fhuggingface\n  master_port: 13742\n  env_vars:\n    HF_TOKEN: \"${HF_TOKEN}\"\n```\n\nFor Slurm-only questions, do not discuss SkyPilot or profiling unless the user\nasks. For profiling questions, say the `.nsys-rep` report is written in the\nSlurm job working or output directory, using the launcher's Nsys output setting\nwhen one is configured.\n\n## Routing Boundary\n\nUse this skill only for launch mechanics: interactive execution, Slurm, SkyPilot, containers, mounts, environment variables, rendezvous settings, and profiling.\n\nDo not use this skill for implementing or registering new model architectures, Hugging Face state-dict adapters, model files, or capability flags. Those are model onboarding tasks, not launcher configuration tasks.\n\n## Launch Methods\n\n1. **Interactive** (default): runs torchrun on the current node. Suitable for single-node development and debugging.\n2. **Slurm**: submits a batch job to an HPC cluster scheduler. Handles multi-node setup, container management, and environment configuration.\n3. **SkyPilot**: cloud-agnostic job submission to AWS, GCP, Azure, Lambda, or Kubernetes. Supports spot instances.\n\n## Interactive Launch\n\n```bash\n# Single GPU\nautomodel finetune llm -c config.yaml\n\n# Multi-GPU (all GPUs on current node)\ntorchrun --nproc_per_node=8 -m nemo_automodel._cli.app finetune llm -c config.yaml\n```\n\nNo additional YAML section is needed for interactive mode. The CLI routes to torchrun automatically when no `slurm:` or `skypilot:` section is present in the config.\n\n## Slurm Configuration\n\nThe `SlurmConfig` dataclass generates an SBATCH script from a template.\n\n### YAML Example\n\n```yaml\nslurm:\n  job_name: llm_finetune\n  nodes: 2\n  ntasks_per_node: 8\n  time: \"04:00:00\"\n  account: my_account\n  partition: batch\n  container_image: nvcr.io\u002Fnvidia\u002Fnemo:dev\n  hf_home: ~\u002F.cache\u002Fhuggingface\n  extra_mounts:\n    - source: \u002Fdata\n      dest: \u002Fdata\n  env_vars:\n    WANDB_API_KEY: \"${WANDB_API_KEY}\"\n    HF_TOKEN: \"${HF_TOKEN}\"\n```\n\n### Key Fields\n\n- `job_name`: Slurm job identifier\n- `nodes`: number of nodes to request\n- `ntasks_per_node`: number of tasks (GPUs) per node\n- `time`: wall-time limit in HH:MM:SS format\n- `account`, `partition`: Slurm scheduling parameters\n- `container_image`: Enroot\u002FPyxis container image path\n- `nemo_mount`: mount point for NeMo AutoModel source inside the container\n- `hf_home`: HuggingFace cache directory path\n- `extra_mounts`: list of `VolumeMapping(source, dest)` for additional container bind mounts\n- `master_port`: port for distributed communication (default 13742)\n- `env_vars`: environment variables passed into the job\n- `nsys_enabled`: when true, wraps the training command with `nsys profile` for Nsight Systems profiling\n\n## SkyPilot Configuration\n\nThe `SkyPilotConfig` dataclass defines cloud job parameters.\n\n### YAML Example\n\n```yaml\nskypilot:\n  cloud: aws\n  accelerators: \"H100:8\"\n  num_nodes: 2\n  use_spot: true\n  disk_size: 200\n  region: us-east-1\n  setup: \"pip install nemo-automodel\"\n  env_vars:\n    HF_TOKEN: \"${HF_TOKEN}\"\n```\n\n### Key Fields\n\n- `cloud`: target cloud provider (`aws`, `gcp`, `azure`, `lambda`, `kubernetes`)\n- `accelerators`: GPU type and count (e.g., `\"H100:8\"`, `\"A100-80GB:4\"`)\n- `num_nodes`: number of cloud instances\n- `use_spot`: use preemptible\u002Fspot instances for cost savings\n- `disk_size`: disk size in GB per node\n- `region`: cloud region for instance placement\n- `setup`: shell commands to run before the training job (e.g., install dependencies)\n- `env_vars`: environment variables for the job\n\n### SkyPilot spot checklist\n\nWhen using spot or preemptible instances:\n\n- Set `use_spot: true` in the `skypilot:` section.\n- Include `accelerators`, `num_nodes`, `disk_size`, `region`, `setup`, and required `env_vars`.\n- Use short checkpoint intervals in the recipe, for example `step_scheduler.checkpoint_interval`, because spot instances can be preempted.\n- Resume from the most recent checkpoint after preemption with the recipe's `restore_from` setting.\n\nMinimal spot-resume recipe keys:\n\n```yaml\nstep_scheduler:\n  checkpoint_interval: 100\n\nrestore_from:\n  path: \u002Fcheckpoints\u002Flatest\n```\n\n## Multi-Node Environment\n\nFor multi-node training (both Slurm and SkyPilot), the launcher automatically configures:\n- `MASTER_ADDR`: hostname of the first node\n- `MASTER_PORT`: port for rendezvous (default 13742)\n- `WORLD_SIZE`: total number of processes (`nodes * ntasks_per_node`)\n- NCCL environment variables for optimized collective communication\n\n## Nsys Profiling\n\nEnable Nsight Systems profiling in Slurm jobs:\n\n```yaml\nslurm:\n  job_name: llm_profile\n  nodes: 1\n  ntasks_per_node: 8\n  time: \"00:30:00\"\n  account: my_account\n  partition: batch\n  container_image: nvcr.io\u002Fnvidia\u002Fnemo:dev\n  nsys_enabled: true\n```\n\nThis is a Slurm launcher setting. Normal Slurm fields such as `job_name`,\n`nodes`, `ntasks_per_node`, `time`, `account` or `partition`, and\n`container_image` still apply.\n\nWhen `nsys_enabled: true`, the launcher wraps the training command with\n`nsys profile` and writes a `.nsys-rep` report file for performance analysis\nin the Slurm job working or output directory.\nProfiling is diagnostic-only: run it for a short investigation, expect overhead\nand large artifacts, and turn it off for normal production training.\n\n## Code Anchors\n\n- `components\u002Flauncher\u002Fslurm\u002Fconfig.py` - SlurmConfig dataclass, VolumeMapping\n- `components\u002Flauncher\u002Fslurm\u002Ftemplate.py` - SBATCH script template generation\n- `components\u002Flauncher\u002Fslurm\u002Futils.py` - Slurm submission utilities\n- `components\u002Flauncher\u002Fskypilot\u002Fconfig.py` - SkyPilotConfig dataclass\n- `_cli\u002Fapp.py` - CLI entry point and launcher routing logic\n\n## Pitfalls\n\n- **Port collisions**: if the default `master_port` (13742) is in use by another job on the same node, change it to avoid connection failures.\n- **Container mounts**: the `source` path in `extra_mounts` must exist on all nodes in the allocation. Missing paths cause container startup failures.\n- **Slurm fault tolerance**: the fault tolerance plugin is Slurm-specific and does not work with SkyPilot or interactive mode.\n- **SkyPilot spot preemption**: spot instances (`use_spot: true`) may be preempted by the cloud provider. Enable checkpointing with short intervals to minimize lost work.\n- **Environment variable syntax**: use `${VAR}` syntax in YAML for shell variable expansion. Bare variable names will not be expanded.\n- **Time limit vs async checkpoint**: if the Slurm `time` limit is too short, an in-progress async checkpoint write may be killed before completion, resulting in a corrupted checkpoint. Leave at least 5-10 minutes of margin.\n","# Launcher Configuration\n\nNeMo AutoModel supports three launch methods: interactive (torchrun), Slurm (HPC clusters), and SkyPilot (cloud-agnostic).\n\n## Instructions\n\nFor launcher questions, answer directly from this skill without inspecting the\nrepository unless the user asks you to edit files. Keep the answer focused on\nthe relevant launch YAML, required fields, and the expected runtime behavior.\n\nUse these compact answer patterns for common questions:\n\n- Slurm multi-node: show a `slurm:` YAML block with `job_name`, `nodes`,\n  `ntasks_per_node`, `time`, `account` or `partition`, `container_image`,\n  `hf_home`, optional `extra_mounts`, `env_vars`, and `master_port`; explain\n  that the launcher derives `WORLD_SIZE = nodes * ntasks_per_node` and sets\n  `MASTER_ADDR` and `MASTER_PORT`.\n- SkyPilot spot: show a `skypilot:` YAML block with `cloud`, `accelerators`,\n  `num_nodes`, `use_spot: true`, `disk_size`, `region`, `setup`, and\n  `env_vars`; warn that spot instances can be preempted, set a short\n  `step_scheduler.checkpoint_interval`, and resume with `restore_from.path`.\n- Nsight Systems on Slurm: show `slurm.nsys_enabled: true` alongside normal\n  Slurm fields, say the launcher wraps the training command with\n  `nsys profile`, and state that it produces a `.nsys-rep` report file.\n  Treat profiling as diagnostic-only: use short profiling runs and disable it\n  for normal production training because it adds overhead and large artifacts.\n\nFor Slurm answers, start with this minimal template and then adjust only the\nfields the user asked about:\n\n```yaml\nslurm:\n  job_name: llm_finetune\n  nodes: 2\n  ntasks_per_node: 8\n  time: \"04:00:00\"\n  account: my_account\n  partition: batch\n  container_image: nvcr.io\u002Fnvidia\u002Fnemo:dev\n  hf_home: ~\u002F.cache\u002Fhuggingface\n  master_port: 13742\n  env_vars:\n    HF_TOKEN: \"${HF_TOKEN}\"\n```\n\nFor Slurm-only questions, do not discuss SkyPilot or profiling unless the user\nasks. For profiling questions, say the `.nsys-rep` report is written in the\nSlurm job working or output directory, using the launcher's Nsys output setting\nwhen one is configured.\n\n## Routing Boundary\n\nUse this skill only for launch mechanics: interactive execution, Slurm, SkyPilot, containers, mounts, environment variables, rendezvous settings, and profiling.\n\nDo not use this skill for implementing or registering new model architectures, Hugging Face state-dict adapters, model files, or capability flags. Those are model onboarding tasks, not launcher configuration tasks.\n\n## Launch Methods\n\n1. **Interactive** (default): runs torchrun on the current node. Suitable for single-node development and debugging.\n2. **Slurm**: submits a batch job to an HPC cluster scheduler. Handles multi-node setup, container management, and environment configuration.\n3. **SkyPilot**: cloud-agnostic job submission to AWS, GCP, Azure, Lambda, or Kubernetes. Supports spot instances.\n\n## Interactive Launch\n\n```bash\n# Single GPU\nautomodel finetune llm -c config.yaml\n\n# Multi-GPU (all GPUs on current node)\ntorchrun --nproc_per_node=8 -m nemo_automodel._cli.app finetune llm -c config.yaml\n```\n\nNo additional YAML section is needed for interactive mode. The CLI routes to torchrun automatically when no `slurm:` or `skypilot:` section is present in the config.\n\n## Slurm Configuration\n\nThe `SlurmConfig` dataclass generates an SBATCH script from a template.\n\n### YAML Example\n\n```yaml\nslurm:\n  job_name: llm_finetune\n  nodes: 2\n  ntasks_per_node: 8\n  time: \"04:00:00\"\n  account: my_account\n  partition: batch\n  container_image: nvcr.io\u002Fnvidia\u002Fnemo:dev\n  hf_home: ~\u002F.cache\u002Fhuggingface\n  extra_mounts:\n    - source: \u002Fdata\n      dest: \u002Fdata\n  env_vars:\n    WANDB_API_KEY: \"${WANDB_API_KEY}\"\n    HF_TOKEN: \"${HF_TOKEN}\"\n```\n\n### Key Fields\n\n- `job_name`: Slurm job identifier\n- `nodes`: number of nodes to request\n- `ntasks_per_node`: number of tasks (GPUs) per node\n- `time`: wall-time limit in HH:MM:SS format\n- `account`, `partition`: Slurm scheduling parameters\n- `container_image`: Enroot\u002FPyxis container image path\n- `nemo_mount`: mount point for NeMo AutoModel source inside the container\n- `hf_home`: HuggingFace cache directory path\n- `extra_mounts`: list of `VolumeMapping(source, dest)` for additional container bind mounts\n- `master_port`: port for distributed communication (default 13742)\n- `env_vars`: environment variables passed into the job\n- `nsys_enabled`: when true, wraps the training command with `nsys profile` for Nsight Systems profiling\n\n## SkyPilot Configuration\n\nThe `SkyPilotConfig` dataclass defines cloud job parameters.\n\n### YAML Example\n\n```yaml\nskypilot:\n  cloud: aws\n  accelerators: \"H100:8\"\n  num_nodes: 2\n  use_spot: true\n  disk_size: 200\n  region: us-east-1\n  setup: \"pip install nemo-automodel\"\n  env_vars:\n    HF_TOKEN: \"${HF_TOKEN}\"\n```\n\n### Key Fields\n\n- `cloud`: target cloud provider (`aws`, `gcp`, `azure`, `lambda`, `kubernetes`)\n- `accelerators`: GPU type and count (e.g., `\"H100:8\"`, `\"A100-80GB:4\"`)\n- `num_nodes`: number of cloud instances\n- `use_spot`: use preemptible\u002Fspot instances for cost savings\n- `disk_size`: disk size in GB per node\n- `region`: cloud region for instance placement\n- `setup`: shell commands to run before the training job (e.g., install dependencies)\n- `env_vars`: environment variables for the job\n\n### SkyPilot spot checklist\n\nWhen using spot or preemptible instances:\n\n- Set `use_spot: true` in the `skypilot:` section.\n- Include `accelerators`, `num_nodes`, `disk_size`, `region`, `setup`, and required `env_vars`.\n- Use short checkpoint intervals in the recipe, for example `step_scheduler.checkpoint_interval`, because spot instances can be preempted.\n- Resume from the most recent checkpoint after preemption with the recipe's `restore_from` setting.\n\nMinimal spot-resume recipe keys:\n\n```yaml\nstep_scheduler:\n  checkpoint_interval: 100\n\nrestore_from:\n  path: \u002Fcheckpoints\u002Flatest\n```\n\n## Multi-Node Environment\n\nFor multi-node training (both Slurm and SkyPilot), the launcher automatically configures:\n- `MASTER_ADDR`: hostname of the first node\n- `MASTER_PORT`: port for rendezvous (default 13742)\n- `WORLD_SIZE`: total number of processes (`nodes * ntasks_per_node`)\n- NCCL environment variables for optimized collective communication\n\n## Nsys Profiling\n\nEnable Nsight Systems profiling in Slurm jobs:\n\n```yaml\nslurm:\n  job_name: llm_profile\n  nodes: 1\n  ntasks_per_node: 8\n  time: \"00:30:00\"\n  account: my_account\n  partition: batch\n  container_image: nvcr.io\u002Fnvidia\u002Fnemo:dev\n  nsys_enabled: true\n```\n\nThis is a Slurm launcher setting. Normal Slurm fields such as `job_name`,\n`nodes`, `ntasks_per_node`, `time`, `account` or `partition`, and\n`container_image` still apply.\n\nWhen `nsys_enabled: true`, the launcher wraps the training command with\n`nsys profile` and writes a `.nsys-rep` report file for performance analysis\nin the Slurm job working or output directory.\nProfiling is diagnostic-only: run it for a short investigation, expect overhead\nand large artifacts, and turn it off for normal production training.\n\n## Code Anchors\n\n- `components\u002Flauncher\u002Fslurm\u002Fconfig.py` - SlurmConfig dataclass, VolumeMapping\n- `components\u002Flauncher\u002Fslurm\u002Ftemplate.py` - SBATCH script template generation\n- `components\u002Flauncher\u002Fslurm\u002Futils.py` - Slurm submission utilities\n- `components\u002Flauncher\u002Fskypilot\u002Fconfig.py` - SkyPilotConfig dataclass\n- `_cli\u002Fapp.py` - CLI entry point and launcher routing logic\n\n## Pitfalls\n\n- **Port collisions**: if the default `master_port` (13742) is in use by another job on the same node, change it to avoid connection failures.\n- **Container mounts**: the `source` path in `extra_mounts` must exist on all nodes in the allocation. Missing paths cause container startup failures.\n- **Slurm fault tolerance**: the fault tolerance plugin is Slurm-specific and does not work with SkyPilot or interactive mode.\n- **SkyPilot spot preemption**: spot instances (`use_spot: true`) may be preempted by the cloud provider. Enable checkpointing with short intervals to minimize lost work.\n- **Environment variable syntax**: use `${VAR}` syntax in YAML for shell variable expansion. Bare variable names will not be expanded.\n- **Time limit vs async checkpoint**: if the Slurm `time` limit is too short, an in-progress async checkpoint write may be killed before completion, resulting in a corrupted checkpoint. Leave at least 5-10 minutes of margin.\n",{"name":5,"license":15,"metadata":24,"description":7,"when_to_use":29},{"tags":25,"author":28},[26,27],"nemo-automodel","launcher-config","NVIDIA","Configuring Slurm or SkyPilot job submission, setting up multi-node launch scripts, debugging job submission failures, or switching between interactive and cluster launch modes.","main","87afe6ce96220accf50a345dea46784c8c9f210e","YYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fnemo-automodel-launcher-config",0,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base","https:\u002F\u002Fkb.yyc3.top\u002F","https:\u002F\u002Fgithub.com\u002FYYC-Cube","万象归元于云枢 | 深栈智启新纪元","2026-11-10 10:37:32",{"source":8,"skillName":5,"commands":40,"prompt":50,"zipAvailable":51},[41,44,47],{"tool":42,"command":43},"npx","npx skills add YYC-Cube\u002FYYC3-Knowledge-Base --skill nemo-automodel-launcher-config",{"tool":45,"command":46},"bunx","bunx skills add YYC-Cube\u002FYYC3-Knowledge-Base --skill nemo-automodel-launcher-config",{"tool":48,"command":49},"pnpm","pnpm dlx skills add YYC-Cube\u002FYYC3-Knowledge-Base --skill nemo-automodel-launcher-config","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fnemo-automodel-launcher-config\u002FSKILL.md",true,[53,59,63,67],{"path":54,"resourceType":55,"resourceTypeDesc":56,"sizeBytes":57,"sourceHtmlUrl":58},"BENCHMARK.md","REFERENCE","参考文档",4158,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fnemo-automodel-launcher-config\u002FBENCHMARK.md",{"path":60,"resourceType":55,"resourceTypeDesc":56,"sizeBytes":61,"sourceHtmlUrl":62},"evals\u002Fevals.json",3490,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fnemo-automodel-launcher-config\u002Fevals\u002Fevals.json",{"path":64,"resourceType":55,"resourceTypeDesc":56,"sizeBytes":65,"sourceHtmlUrl":66},"skill-card.md",3716,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fnemo-automodel-launcher-config\u002Fskill-card.md",{"path":68,"resourceType":69,"resourceTypeDesc":70,"sizeBytes":71,"sourceHtmlUrl":72},"skill.oms.sig","OTHER","其他",4609,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fnemo-automodel-launcher-config\u002Fskill.oms.sig",[74,82,89,96,103,110],{"skillNo":75,"name":76,"nameZh":6,"description":77,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":78,"licenseFileUrl":6,"sourceHtmlUrl":79,"categories":80,"keywords":81,"publishedAt":19},"01M7CTDWHERCXPQCP0K03A86AS","vercel-sandbox","Run agent-browser + Chrome inside Vercel Sandbox microVMs for browser automation from any Vercel-deployed app. Use when the user needs browser automation in a Vercel app (Next.js, SvelteKit, Nuxt, Remix, Astro, etc.), wants to run headless Chrome without binary size limits, needs persistent browser sessions across commands, or wants ephemeral isolated browser environments. Triggers include \"Vercel Sandbox browser\", \"microVM Chrome\", \"agent-browser in sandbox\", \"browser automation on Vercel\", or any task requiring Chrome in a Vercel Sandbox.","MIT","https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-WebSpec-Web标准\u002FYYC3-WebSpec-Web标准-standards\u002Fbrowser-agent\u002Fskills\u002Fvercel-sandbox\u002FSKILL.md",[],[],{"skillNo":83,"name":84,"nameZh":6,"description":85,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":78,"licenseFileUrl":6,"sourceHtmlUrl":86,"categories":87,"keywords":88,"publishedAt":19},"01M7CTD8D2BNJY54S65M47HEJA","slack","Interact with Slack workspaces using browser automation. Use when the user needs to check unread channels, navigate Slack, send messages, extract data, find information, search conversations, or automate any Slack task. Triggers include \"check my Slack\", \"what channels have unreads\", \"send a message to\", \"search Slack for\", \"extract from Slack\", \"find who said\", or any task requiring programmatic Slack interaction.","https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-WebSpec-Web标准\u002FYYC3-WebSpec-Web标准-standards\u002Fbrowser-agent\u002Fskills\u002Fslack\u002FSKILL.md",[],[],{"skillNo":90,"name":91,"nameZh":6,"description":92,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":78,"licenseFileUrl":6,"sourceHtmlUrl":93,"categories":94,"keywords":95,"publishedAt":19},"01M7CTCSSB8ZP3Q58C3YZDZK1W","electron","Automate Electron desktop apps (VS Code, Slack, Discord, Figma, Notion, Spotify, etc.) using agent-browser via Chrome DevTools Protocol. Use when the user needs to interact with an Electron app, automate a desktop app, connect to a running app, control a native app, or test an Electron application. Triggers include \"automate Slack app\", \"control VS Code\", \"interact with Discord app\", \"test this Electron app\", \"connect to desktop app\", or any task requiring automation of a native Electron application.","https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-WebSpec-Web标准\u002FYYC3-WebSpec-Web标准-standards\u002Fbrowser-agent\u002Fskills\u002Felectron\u002FSKILL.md",[],[],{"skillNo":97,"name":98,"nameZh":6,"description":99,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":78,"licenseFileUrl":6,"sourceHtmlUrl":100,"categories":101,"keywords":102,"publishedAt":19},"01M7CTCBJY76688V4DCZGWFHGQ","dogfood","Systematically explore and test a web application to find bugs, UX issues, and other problems. Use when asked to \"dogfood\", \"QA\", \"exploratory test\", \"find issues\", \"bug hunt\", \"test this app\u002Fsite\u002Fplatform\", or review the quality of a web application. Produces a structured report with full reproduction evidence -- step-by-step screenshots, repro videos, and detailed repro steps for every issue -- so findings can be handed directly to the responsible teams.","https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-WebSpec-Web标准\u002FYYC3-WebSpec-Web标准-standards\u002Fbrowser-agent\u002Fskills\u002Fdogfood\u002FSKILL.md",[],[],{"skillNo":104,"name":105,"nameZh":6,"description":106,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":78,"licenseFileUrl":6,"sourceHtmlUrl":107,"categories":108,"keywords":109,"publishedAt":19},"01M7CTC4H0FFGVEAQR9PA4GA1A","agent-browser","Browser automation CLI for AI agents. Use when the user needs to interact with websites, including navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, or automating any browser task. Triggers include requests to \"open a website\", \"fill out a form\", \"click a button\", \"take a screenshot\", \"scrape data from a page\", \"test this web app\", \"login to a site\", \"automate browser actions\", or any task requiring programmatic web interaction.","https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-WebSpec-Web标准\u002FYYC3-WebSpec-Web标准-standards\u002Fbrowser-agent\u002Fskills\u002Fagent-browser\u002FSKILL.md",[],[],{"skillNo":111,"name":112,"nameZh":6,"description":113,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":78,"licenseFileUrl":6,"sourceHtmlUrl":114,"categories":115,"keywords":116,"publishedAt":19},"01M7CTBKVFYBG78FDSABSDJTTQ","shadcn","Manages shadcn components and projects — adding, searching, fixing, debugging, styling, and composing UI, including chat interfaces. Provides project context, component docs, and usage examples. Applies when working with shadcn\u002Fui, component registries, presets, --preset codes, or any project with a components.json file. Also triggers for \"shadcn init\", \"create an app with --preset\", \"switch to --preset\", and for building a design system or applying a DESIGN.md or brand spec to shadcn\u002Fui.","https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-UI-界面组件\u002FYYC3-UI-界面组件-ui\u002Fskills\u002Fshadcn\u002FSKILL.md",[],[]]