[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6F8DZJRZ5DFXK7KJR2YTF46":3,"skill-same:01M6F8DZJRZ5DFXK7KJR2YTF46":79},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":46,"linkOnly":47,"markdownContent":48,"markdownBody":49,"frontMatter":50,"refName":33,"commitSha":59,"skillDirectory":60,"forksCount":61,"repositoryHtmlUrl":62,"repositoryHomepageUrl":6,"ownerHtmlUrl":63,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":64,"install":65,"resources":78},"01M6F8DZJRZ5DFXK7KJR2YTF46","llm-inference-scaling",null,"Auto-scale LLM inference clusters on Kubernetes using KEDA, custom GPU metrics, and horizontal pod autoscaling.","boisenoise\u002Fskills-collections","boisenoise","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F193180?v=4","01M6CDBDQK6BDN55ZMA6B6PKBM","01M6CDBDVSJ3JPTBYN26FZ7TZK",12,"MIT","https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections\u002Fblob\u002Fmain\u002Fskills\u002Fantigravity-llm-inference-scaling\u002FSKILL.md",[],[18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45],"inference","llm inference scaling","llm inference","scaling","community","Clusters","2026","autoscaling","keda","relevant","helper","scripts","critical","os","platform","main","github","agent-skills","only","auto","devops","access","devops-security","gpu","license","metrics","security","kubernetes","2026-11-10 10:12:30",false,"---\nname: llm-inference-scaling\ndescription: Auto-scale LLM inference clusters on Kubernetes using KEDA, custom GPU\n  metrics, and horizontal pod autoscaling.\ncategory: devops\nrisk: critical\nsource: https:\u002F\u002Fgithub.com\u002FBagelHole\u002FDevOps-Security-Agent-Skills\nsource_repo: BagelHole\u002FDevOps-Security-Agent-Skills\nsource_type: community\ndate_added: '2026-09-20'\nlicense: MIT\nlicense_source: https:\u002F\u002Fgithub.com\u002FBagelHole\u002FDevOps-Security-Agent-Skills\u002Fblob\u002Fmain\u002FLICENSE\ncompatibility: Requires the relevant OS\u002Fplatform tooling and privileged access where\n  noted. Docs-only; helper scripts and templates not bundled.\nmetadata:\n  author: devops-skills\n  version: '1.0'\n---\n\n# LLM Inference Scaling\n\nScale LLM inference horizontally on Kubernetes with GPU-aware autoscaling, request queuing, and cost-efficient spot instance strategies.\n\n## When to Use This Skill\n\nUse this skill when:\n- LLM API traffic is unpredictable and you need to scale up\u002Fdown automatically\n- Managing a fleet of vLLM or TGI inference pods on Kubernetes\n- Reducing inference costs with spot\u002Fpreemptible GPU instances\n- Implementing queue-based autoscaling for batch inference jobs\n- Building a multi-model serving platform that shares GPU resources\n\n## Prerequisites\n\n- Kubernetes cluster with GPU nodes (NVIDIA operator installed)\n- KEDA (Kubernetes Event-Driven Autoscaler) installed\n- Prometheus with GPU metrics (`dcgm-exporter` or `gpu-operator`)\n- Helm 3+ for chart deployments\n\n## GPU Node Setup\n\n```bash\n# Install NVIDIA GPU Operator (handles drivers, container toolkit, DCGM)\nhelm repo add nvidia https:\u002F\u002Fhelm.ngc.nvidia.com\u002Fnvidia\nhelm repo update\n\nhelm install gpu-operator nvidia\u002Fgpu-operator \\\n  --namespace gpu-operator \\\n  --create-namespace \\\n  --set driver.enabled=true \\\n  --set dcgm.enabled=true \\\n  --set devicePlugin.enabled=true\n\n# Verify GPU nodes are recognized\nkubectl get nodes -l nvidia.com\u002Fgpu.present=true\nkubectl describe node \u003Cgpu-node> | grep nvidia\n```\n\n## vLLM Deployment with GPU Resources\n\n```yaml\napiVersion: apps\u002Fv1\nkind: Deployment\nmetadata:\n  name: vllm-llama-8b\n  labels:\n    app: vllm\n    model: llama-3.1-8b\nspec:\n  replicas: 1\n  selector:\n    matchLabels:\n      app: vllm\n      model: llama-3.1-8b\n  template:\n    metadata:\n      labels:\n        app: vllm\n        model: llama-3.1-8b\n    spec:\n      nodeSelector:\n        nvidia.com\u002Fgpu.present: \"true\"\n      tolerations:\n      - key: nvidia.com\u002Fgpu\n        operator: Exists\n        effect: NoSchedule\n      containers:\n      - name: vllm\n        image: vllm\u002Fvllm-openai:latest\n        args:\n        - \"--model\"\n        - \"meta-llama\u002FLlama-3.1-8B-Instruct\"\n        - \"--tensor-parallel-size\"\n        - \"1\"\n        - \"--gpu-memory-utilization\"\n        - \"0.90\"\n        - \"--max-num-seqs\"\n        - \"128\"\n        resources:\n          requests:\n            nvidia.com\u002Fgpu: \"1\"\n            memory: \"20Gi\"\n            cpu: \"4\"\n          limits:\n            nvidia.com\u002Fgpu: \"1\"\n            memory: \"24Gi\"\n            cpu: \"8\"\n        ports:\n        - containerPort: 8000\n        readinessProbe:\n          httpGet:\n            path: \u002Fhealth\n            port: 8000\n          initialDelaySeconds: 60\n          periodSeconds: 10\n        env:\n        - name: HUGGING_FACE_HUB_TOKEN\n          valueFrom:\n            secretKeyRef:\n              name: hf-token\n              key: token\n```\n\n## KEDA Autoscaling on Prometheus Metrics\n\n```yaml\napiVersion: keda.sh\u002Fv1alpha1\nkind: ScaledObject\nmetadata:\n  name: vllm-scaledobject\nspec:\n  scaleTargetRef:\n    name: vllm-llama-8b\n  minReplicaCount: 1\n  maxReplicaCount: 8\n  cooldownPeriod: 300          # 5 min before scale-down\n  pollingInterval: 15\n  triggers:\n  - type: prometheus\n    metadata:\n      serverAddress: http:\u002F\u002Fprometheus-server.monitoring:9090\n      metricName: vllm_num_requests_waiting\n      threshold: \"10\"           # scale up if >10 requests waiting\n      query: |\n        sum(vllm:num_requests_waiting{deployment=\"vllm-llama-8b\"})\n  - type: prometheus\n    metadata:\n      serverAddress: http:\u002F\u002Fprometheus-server.monitoring:9090\n      metricName: vllm_gpu_cache_usage\n      threshold: \"0.8\"          # scale up if KV cache >80% full\n      query: |\n        avg(vllm:gpu_cache_usage_perc{deployment=\"vllm-llama-8b\"})\n```\n\n## Queue-Based Scaling (Redis + KEDA)\n\n```yaml\n# ScaledJob for async batch inference\napiVersion: keda.sh\u002Fv1alpha1\nkind: ScaledJob\nmetadata:\n  name: llm-batch-inference\nspec:\n  jobTargetRef:\n    template:\n      spec:\n        containers:\n        - name: inference-worker\n          image: myapp\u002Finference-worker:latest\n          env:\n          - name: REDIS_URL\n            value: redis:\u002F\u002Fredis:6379\n          - name: QUEUE_NAME\n            value: inference-jobs\n        restartPolicy: OnFailure\n  minReplicaCount: 0\n  maxReplicaCount: 20\n  pollingInterval: 5\n  successfulJobsHistoryLimit: 3\n  triggers:\n  - type: redis\n    metadata:\n      address: redis:6379\n      listName: inference-jobs\n      listLength: \"5\"           # 1 worker per 5 queued jobs\n```\n\n## Spot Instance Strategy\n\n```yaml\n# Mixed node pool: on-demand + spot GPUs\napiVersion: v1\nkind: ConfigMap\nmetadata:\n  name: cluster-autoscaler-priority-config\ndata:\n  priorities: |\n    10:  # lower priority; higher numeric priority is preferred\n    - .*spot.*\n    50:\n    - .*on-demand.*\n---\n# Node affinity for spot with on-demand fallback\nspec:\n  affinity:\n    nodeAffinity:\n      preferredDuringSchedulingIgnoredDuringExecution:\n      - weight: 80\n        preference:\n          matchExpressions:\n          - key: node.kubernetes.io\u002Flifecycle\n            operator: In\n            values: [spot]\n      - weight: 20\n        preference:\n          matchExpressions:\n          - key: node.kubernetes.io\u002Flifecycle\n            operator: In\n            values: [on-demand]\n```\n\n## Cluster Autoscaler for GPU Nodes\n\n```bash\n# AWS EKS — enable cluster autoscaler for GPU node group\nhelm install cluster-autoscaler autoscaler\u002Fcluster-autoscaler \\\n  --namespace kube-system \\\n  --set autoDiscovery.clusterName=my-cluster \\\n  --set awsRegion=us-east-1 \\\n  --set rbac.serviceAccount.annotations.\"eks\\.amazonaws\\.com\u002Frole-arn\"=arn:aws:iam::ACCOUNT:role\u002FClusterAutoscalerRole \\\n  --set extraArgs.skip-nodes-with-local-storage=false \\\n  --set extraArgs.expander=least-waste\n\n# Annotate GPU node group for autoscaler\nkubectl annotate node \u003Cnode> \\\n  cluster-autoscaler.kubernetes.io\u002Fsafe-to-evict=\"false\"\n```\n\n## Scaling Metrics to Monitor\n\n```bash\n# Prometheus queries for scaling decisions\n# Requests waiting in vLLM queue\nsum(vllm:num_requests_waiting) by (model)\n\n# GPU KV cache utilization (>80% = bottleneck)\navg(vllm:gpu_cache_usage_perc) by (pod)\n\n# Tokens per second throughput\nsum(rate(vllm:generation_tokens_total[5m])) by (model)\n\n# P99 time-to-first-token\nhistogram_quantile(0.99, rate(vllm:time_to_first_token_seconds_bucket[5m]))\n```\n\n## Common Issues\n\n| Issue | Cause | Fix |\n|-------|-------|-----|\n| Pods stuck in `Pending` | No GPU nodes available | Check cluster autoscaler logs; verify node group limits |\n| Scale-up too slow | Cluster autoscaler delay + model load time | Pre-warm replicas; increase `minReplicaCount` |\n| GPU fragmentation | Multiple small models on large GPUs | Use MIG partitioning or consolidate model sizes |\n| Spot eviction causes errors | Spot instance reclamation | Add `PodDisruptionBudget`; use graceful shutdown |\n| KEDA not scaling | Prometheus query returns no data | Test query in Prometheus UI first |\n\n## Best Practices\n\n- Set `minReplicaCount: 1` to avoid cold starts; scale to 0 only for batch jobs.\n- Use `PodDisruptionBudget` with `minAvailable: 1` to survive spot evictions.\n- Pre-pull model weights into a shared PVC to potentially reduce model download time; measure startup on the actual cluster.\n- Separate model families across node pools (A10G for 7B, A100 for 70B).\n- Use Kubernetes VPA for CPU\u002Fmemory right-sizing alongside KEDA for replica count.\n\n## Related Skills\n\n- vllm-server (`vllm-server`) - vLLM configuration and tuning\n- gpu-server-management (`gpu-server-management`) - GPU node setup\n- model-serving-kubernetes (`model-serving-kubernetes`) - KServe\n- kubernetes-ops (`kubernetes-ops`) - Core Kubernetes\n- llm-cost-optimization (`llm-cost-optimization`) - Cost strategies\n\n## Limitations\n\n- Infrastructure commands can disrupt services: confirm target host\u002Fscope and have backups\u002Fsnapshots before mutating state.\n- Docs-only import: upstream scripts and templates not bundled.\n\nPodDisruptionBudgets constrain voluntary disruptions, not spot preemption or node failure. GPU workers need explicit `nvidia.com\u002Fgpu` resource requests\u002Flimits and a compatible device plugin; a ScaledJob example without them is incomplete. Aggregate classic Prometheus histogram buckets with `sum by (le)` before `histogram_quantile`. Pin reviewed images, validate metric labels\u002Fscaler schemas and bound cost before autoscaling.\n","# LLM Inference Scaling\n\nScale LLM inference horizontally on Kubernetes with GPU-aware autoscaling, request queuing, and cost-efficient spot instance strategies.\n\n## When to Use This Skill\n\nUse this skill when:\n- LLM API traffic is unpredictable and you need to scale up\u002Fdown automatically\n- Managing a fleet of vLLM or TGI inference pods on Kubernetes\n- Reducing inference costs with spot\u002Fpreemptible GPU instances\n- Implementing queue-based autoscaling for batch inference jobs\n- Building a multi-model serving platform that shares GPU resources\n\n## Prerequisites\n\n- Kubernetes cluster with GPU nodes (NVIDIA operator installed)\n- KEDA (Kubernetes Event-Driven Autoscaler) installed\n- Prometheus with GPU metrics (`dcgm-exporter` or `gpu-operator`)\n- Helm 3+ for chart deployments\n\n## GPU Node Setup\n\n```bash\n# Install NVIDIA GPU Operator (handles drivers, container toolkit, DCGM)\nhelm repo add nvidia https:\u002F\u002Fhelm.ngc.nvidia.com\u002Fnvidia\nhelm repo update\n\nhelm install gpu-operator nvidia\u002Fgpu-operator \\\n  --namespace gpu-operator \\\n  --create-namespace \\\n  --set driver.enabled=true \\\n  --set dcgm.enabled=true \\\n  --set devicePlugin.enabled=true\n\n# Verify GPU nodes are recognized\nkubectl get nodes -l nvidia.com\u002Fgpu.present=true\nkubectl describe node \u003Cgpu-node> | grep nvidia\n```\n\n## vLLM Deployment with GPU Resources\n\n```yaml\napiVersion: apps\u002Fv1\nkind: Deployment\nmetadata:\n  name: vllm-llama-8b\n  labels:\n    app: vllm\n    model: llama-3.1-8b\nspec:\n  replicas: 1\n  selector:\n    matchLabels:\n      app: vllm\n      model: llama-3.1-8b\n  template:\n    metadata:\n      labels:\n        app: vllm\n        model: llama-3.1-8b\n    spec:\n      nodeSelector:\n        nvidia.com\u002Fgpu.present: \"true\"\n      tolerations:\n      - key: nvidia.com\u002Fgpu\n        operator: Exists\n        effect: NoSchedule\n      containers:\n      - name: vllm\n        image: vllm\u002Fvllm-openai:latest\n        args:\n        - \"--model\"\n        - \"meta-llama\u002FLlama-3.1-8B-Instruct\"\n        - \"--tensor-parallel-size\"\n        - \"1\"\n        - \"--gpu-memory-utilization\"\n        - \"0.90\"\n        - \"--max-num-seqs\"\n        - \"128\"\n        resources:\n          requests:\n            nvidia.com\u002Fgpu: \"1\"\n            memory: \"20Gi\"\n            cpu: \"4\"\n          limits:\n            nvidia.com\u002Fgpu: \"1\"\n            memory: \"24Gi\"\n            cpu: \"8\"\n        ports:\n        - containerPort: 8000\n        readinessProbe:\n          httpGet:\n            path: \u002Fhealth\n            port: 8000\n          initialDelaySeconds: 60\n          periodSeconds: 10\n        env:\n        - name: HUGGING_FACE_HUB_TOKEN\n          valueFrom:\n            secretKeyRef:\n              name: hf-token\n              key: token\n```\n\n## KEDA Autoscaling on Prometheus Metrics\n\n```yaml\napiVersion: keda.sh\u002Fv1alpha1\nkind: ScaledObject\nmetadata:\n  name: vllm-scaledobject\nspec:\n  scaleTargetRef:\n    name: vllm-llama-8b\n  minReplicaCount: 1\n  maxReplicaCount: 8\n  cooldownPeriod: 300          # 5 min before scale-down\n  pollingInterval: 15\n  triggers:\n  - type: prometheus\n    metadata:\n      serverAddress: http:\u002F\u002Fprometheus-server.monitoring:9090\n      metricName: vllm_num_requests_waiting\n      threshold: \"10\"           # scale up if >10 requests waiting\n      query: |\n        sum(vllm:num_requests_waiting{deployment=\"vllm-llama-8b\"})\n  - type: prometheus\n    metadata:\n      serverAddress: http:\u002F\u002Fprometheus-server.monitoring:9090\n      metricName: vllm_gpu_cache_usage\n      threshold: \"0.8\"          # scale up if KV cache >80% full\n      query: |\n        avg(vllm:gpu_cache_usage_perc{deployment=\"vllm-llama-8b\"})\n```\n\n## Queue-Based Scaling (Redis + KEDA)\n\n```yaml\n# ScaledJob for async batch inference\napiVersion: keda.sh\u002Fv1alpha1\nkind: ScaledJob\nmetadata:\n  name: llm-batch-inference\nspec:\n  jobTargetRef:\n    template:\n      spec:\n        containers:\n        - name: inference-worker\n          image: myapp\u002Finference-worker:latest\n          env:\n          - name: REDIS_URL\n            value: redis:\u002F\u002Fredis:6379\n          - name: QUEUE_NAME\n            value: inference-jobs\n        restartPolicy: OnFailure\n  minReplicaCount: 0\n  maxReplicaCount: 20\n  pollingInterval: 5\n  successfulJobsHistoryLimit: 3\n  triggers:\n  - type: redis\n    metadata:\n      address: redis:6379\n      listName: inference-jobs\n      listLength: \"5\"           # 1 worker per 5 queued jobs\n```\n\n## Spot Instance Strategy\n\n```yaml\n# Mixed node pool: on-demand + spot GPUs\napiVersion: v1\nkind: ConfigMap\nmetadata:\n  name: cluster-autoscaler-priority-config\ndata:\n  priorities: |\n    10:  # lower priority; higher numeric priority is preferred\n    - .*spot.*\n    50:\n    - .*on-demand.*\n---\n# Node affinity for spot with on-demand fallback\nspec:\n  affinity:\n    nodeAffinity:\n      preferredDuringSchedulingIgnoredDuringExecution:\n      - weight: 80\n        preference:\n          matchExpressions:\n          - key: node.kubernetes.io\u002Flifecycle\n            operator: In\n            values: [spot]\n      - weight: 20\n        preference:\n          matchExpressions:\n          - key: node.kubernetes.io\u002Flifecycle\n            operator: In\n            values: [on-demand]\n```\n\n## Cluster Autoscaler for GPU Nodes\n\n```bash\n# AWS EKS — enable cluster autoscaler for GPU node group\nhelm install cluster-autoscaler autoscaler\u002Fcluster-autoscaler \\\n  --namespace kube-system \\\n  --set autoDiscovery.clusterName=my-cluster \\\n  --set awsRegion=us-east-1 \\\n  --set rbac.serviceAccount.annotations.\"eks\\.amazonaws\\.com\u002Frole-arn\"=arn:aws:iam::ACCOUNT:role\u002FClusterAutoscalerRole \\\n  --set extraArgs.skip-nodes-with-local-storage=false \\\n  --set extraArgs.expander=least-waste\n\n# Annotate GPU node group for autoscaler\nkubectl annotate node \u003Cnode> \\\n  cluster-autoscaler.kubernetes.io\u002Fsafe-to-evict=\"false\"\n```\n\n## Scaling Metrics to Monitor\n\n```bash\n# Prometheus queries for scaling decisions\n# Requests waiting in vLLM queue\nsum(vllm:num_requests_waiting) by (model)\n\n# GPU KV cache utilization (>80% = bottleneck)\navg(vllm:gpu_cache_usage_perc) by (pod)\n\n# Tokens per second throughput\nsum(rate(vllm:generation_tokens_total[5m])) by (model)\n\n# P99 time-to-first-token\nhistogram_quantile(0.99, rate(vllm:time_to_first_token_seconds_bucket[5m]))\n```\n\n## Common Issues\n\n| Issue | Cause | Fix |\n|-------|-------|-----|\n| Pods stuck in `Pending` | No GPU nodes available | Check cluster autoscaler logs; verify node group limits |\n| Scale-up too slow | Cluster autoscaler delay + model load time | Pre-warm replicas; increase `minReplicaCount` |\n| GPU fragmentation | Multiple small models on large GPUs | Use MIG partitioning or consolidate model sizes |\n| Spot eviction causes errors | Spot instance reclamation | Add `PodDisruptionBudget`; use graceful shutdown |\n| KEDA not scaling | Prometheus query returns no data | Test query in Prometheus UI first |\n\n## Best Practices\n\n- Set `minReplicaCount: 1` to avoid cold starts; scale to 0 only for batch jobs.\n- Use `PodDisruptionBudget` with `minAvailable: 1` to survive spot evictions.\n- Pre-pull model weights into a shared PVC to potentially reduce model download time; measure startup on the actual cluster.\n- Separate model families across node pools (A10G for 7B, A100 for 70B).\n- Use Kubernetes VPA for CPU\u002Fmemory right-sizing alongside KEDA for replica count.\n\n## Related Skills\n\n- vllm-server (`vllm-server`) - vLLM configuration and tuning\n- gpu-server-management (`gpu-server-management`) - GPU node setup\n- model-serving-kubernetes (`model-serving-kubernetes`) - KServe\n- kubernetes-ops (`kubernetes-ops`) - Core Kubernetes\n- llm-cost-optimization (`llm-cost-optimization`) - Cost strategies\n\n## Limitations\n\n- Infrastructure commands can disrupt services: confirm target host\u002Fscope and have backups\u002Fsnapshots before mutating state.\n- Docs-only import: upstream scripts and templates not bundled.\n\nPodDisruptionBudgets constrain voluntary disruptions, not spot preemption or node failure. GPU workers need explicit `nvidia.com\u002Fgpu` resource requests\u002Flimits and a compatible device plugin; a ScaledJob example without them is incomplete. Aggregate classic Prometheus histogram buckets with `sum by (le)` before `histogram_quantile`. Pin reviewed images, validate metric labels\u002Fscaler schemas and bound cost before autoscaling.\n",{"name":5,"risk":30,"source":51,"license":14,"category":38,"metadata":52,"date_added":55,"description":7,"source_repo":56,"source_type":22,"compatibility":57,"license_source":58},"https:\u002F\u002Fgithub.com\u002FBagelHole\u002FDevOps-Security-Agent-Skills",{"author":53,"version":54},"devops-skills","1.0","2026-09-20","BagelHole\u002FDevOps-Security-Agent-Skills","Requires the relevant OS\u002Fplatform tooling and privileged access where noted. Docs-only; helper scripts and templates not bundled.","https:\u002F\u002Fgithub.com\u002FBagelHole\u002FDevOps-Security-Agent-Skills\u002Fblob\u002Fmain\u002FLICENSE","7bf2279c8a65758319526a721d5c4b75683aec9c","skills\u002Fantigravity-llm-inference-scaling",3,"https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections","https:\u002F\u002Fgithub.com\u002Fboisenoise","2026-11-11 14:20:43",{"source":8,"skillName":5,"commands":66,"prompt":76,"zipAvailable":77},[67,70,73],{"tool":68,"command":69},"npx","npx skills add boisenoise\u002Fskills-collections --skill llm-inference-scaling",{"tool":71,"command":72},"bunx","bunx skills add boisenoise\u002Fskills-collections --skill llm-inference-scaling",{"tool":74,"command":75},"pnpm","pnpm dlx skills add boisenoise\u002Fskills-collections --skill llm-inference-scaling","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections\u002Fblob\u002Fmain\u002Fskills\u002Fantigravity-llm-inference-scaling\u002FSKILL.md",true,[],[80,105,131,157,176,199],{"skillNo":81,"name":82,"nameZh":6,"description":83,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":84,"licenseFileUrl":85,"sourceHtmlUrl":86,"categories":87,"keywords":88,"publishedAt":46},"01M6WNH92BHVFXHGPDZCPJ0AG0","web-access","Use Treg for public web or platform data only when existing search, browser and connected tools cannot cover the task, with explicit cost and data boundaries.","Apache-2.0","https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections\u002Fblob\u002Fmain\u002Fskills\u002Fantigravity-web-access\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections\u002Fblob\u002Fmain\u002Fskills\u002Fantigravity-web-access\u002FSKILL.md",[],[89,39,90,91,92,93,94,95,96,97,98,36,99,100,32,30,101,102,103,104,24],"web","web access","official","development","browser","search","data","Tools","read","fetch","youtube","free tools","free","cover","normal","when",{"skillNo":106,"name":107,"nameZh":6,"description":108,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":84,"licenseFileUrl":109,"sourceHtmlUrl":110,"categories":111,"keywords":117,"publishedAt":46},"01M6WNF9XAYAB2CZDDJBAKNM93","treg","Use the Treg CLI or MCP catalog to discover, price and call authorized external tools with team-scoped credentials and a user-approved budget.","https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections\u002Fblob\u002Fmain\u002Fskills\u002Fantigravity-treg\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections\u002Fblob\u002Fmain\u002Fskills\u002Fantigravity-treg\u002FSKILL.md",[112],{"categoryNo":113,"categoryName":114,"taxonomyType":115,"taxonomyTypeDesc":116},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",[107,118,119,34,91,92,42,96,120,121,122,123,124,125,126,30,127,128,129,130,24],"MCP","cli","Budget","credentials","team","catalog","use","call","discover","external tools","with","scoped","blob",{"skillNo":132,"name":133,"nameZh":6,"description":134,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":84,"licenseFileUrl":135,"sourceHtmlUrl":136,"categories":137,"keywords":143,"publishedAt":46},"01M6WN93MEAZ4E41DYAHJRPHQT","recruit","Turn a job description into an evidence-linked candidate shortlist through Treg, with job-related ranking and optional authorized business contact enrichment.","https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections\u002Fblob\u002Fmain\u002Fskills\u002Fantigravity-recruit\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections\u002Fblob\u002Fmain\u002Fskills\u002Fantigravity-recruit\u002FSKILL.md",[138],{"categoryNo":139,"categoryName":140,"taxonomyType":141,"taxonomyTypeDesc":142},"01JBSEEDCAT000000000000038","人力资源","PROFESSION","适用职业",[133,34,91,42,144,145,146,147,148,149,30,150,151,152,128,153,154,155,156,130,107,24],"https","email","enrichment","business","evidence","phone","ranking","post","job description","for","person","candidate","reason",{"skillNo":158,"name":159,"nameZh":6,"description":160,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":84,"licenseFileUrl":161,"sourceHtmlUrl":162,"categories":163,"keywords":164,"publishedAt":46},"01M6WN8R5RXSMHM9B4D9FAFX42","rea-growth","Analyze a company’s public growth and content through Treg, distinguish observations from hypotheses, and produce an evidence-linked playbook within an approved budget.","https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections\u002Fblob\u002Fmain\u002Fskills\u002Fantigravity-rea-growth\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections\u002Fblob\u002Fmain\u002Fskills\u002Fantigravity-rea-growth\u002FSKILL.md",[],[165,166,167,34,91,42,168,169,144,170,120,171,172,173,148,30,174,175,130,107,24],"growth","rea","rea growth","produce","content","playbook","marketing","analyze","public","observations","company",{"skillNo":177,"name":178,"nameZh":6,"description":179,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":84,"licenseFileUrl":180,"sourceHtmlUrl":181,"categories":182,"keywords":183,"publishedAt":46},"01M6WN2EVE3RQRGWAKBT4S7YT2","make-ugc","Create original AI UGC-style product videos through Treg: research hooks, approve a licensed presenter and voice, generate clips, add captions and report costs.","https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections\u002Fblob\u002Fmain\u002Fskills\u002Fantigravity-make-ugc\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections\u002Fblob\u002Fmain\u002Fskills\u002Fantigravity-make-ugc\u002FSKILL.md",[],[184,185,34,186,91,187,42,188,189,190,191,192,30,193,194,195,196,197,198,130],"make","make ugc","research","create","Video","Captions","ads","short","hooks","head","style","niche","trending","talking-head","approve",{"skillNo":200,"name":201,"nameZh":6,"description":202,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":84,"licenseFileUrl":203,"sourceHtmlUrl":204,"categories":205,"keywords":206,"publishedAt":46},"01M6WN10ARAA63Y8K1YNGXJ2RM","lead-signals","Find and qualify public buying signals through Treg, rank fitting business accounts and optionally enrich authorized business contact information.","https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections\u002Fblob\u002Fmain\u002Fskills\u002Fantigravity-lead-signals\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections\u002Fblob\u002Fmain\u002Fskills\u002Fantigravity-lead-signals\u002FSKILL.md",[],[207,208,209,34,91,42,210,211,147,212,213,30,214,215,216,217,218,219,220,130],"signals","lead","lead signals","how","find","ready","as","how to","qualify","people","information","enrich","changes","funding"]