DevOps-Security-Agent-Skills

Большая база скиллов по DevOps, безопасности, инфраструктуре и комплаенсу для AI-агентов

Скилл

Низкий риск

Низкий уровень ставим, когда запись в основном дает агенту инструкции и справочные материалы.

Почему такой уровень

  • Скиллы состоят из справочников и практик, операции не выполняют
Все причины и проверки

bagelhole/devops-security-agent-skills

Установка

В терминале, через SkillFoxx CLI

npx skillfoxx add skills/devops-security-agent-skills

Определит агенты на машине, проверит риск и зафиксирует версию.

Другие способы установки

Собрано автоматически, проверьте перед установкой.

Выполните в терминале в папке проекта

npx skills add bagelhole/devops-security-agent-skills --skill audit-logging aws-cloudtrail azure-monitor-audit gcp-audit-logs business-continuity disaster-recovery incident-management runbook-creation fedramp-compliance gdpr-compliance hipaa-compliance iso27001-compliance pci-dss-compliance soc2-compliance access-review asset-inventory change-management policy-as-code vendor-management agent-evals agent-observability ai-pipeline-orchestration ai-sre-incident-response llm-caching llm-cost-optimization llmops-platform-engineering model-registry-governance rag-observability-evals azure-devops circleci github-actions gitlab-ci jenkins container-registries docker-compose docker-management podman devcontainers-nix alerting-oncall datadog ebpf-observability elk-stack loki-logging new-relic opentelemetry prometheus-grafana sre-dashboards argocd-gitops helm-charts kubernetes-ops kustomize model-serving-kubernetes openshift platform-engineering blue-green-deploy feature-flags git-workflow semantic-versioning aws-cost-optimization aws-ec2 aws-ecs-fargate aws-iam aws-lambda aws-rds aws-s3 aws-vpc cloudformation terraform-aws arm-templates azure-aks azure-functions azure-networking azure-sql azure-vms terraform-azure gcp-cloud-functions gcp-cloud-sql gcp-compute gcp-gke gcp-networking terraform-gcp cloudflare-pages cloudflare-r2 cloudflare-workers cloudflare-zero-trust database-backups mongodb mysql planetscale postgresql redis vector-database-ops opentofu-migration identity-access-management mdm-device-management saas-security-posture startup-it-troubleshooting gpu-kubernetes-operations llm-fine-tuning llm-inference-scaling mac-mini-llm-lab multi-tenant-llm-hosting ollama-stack openclaw-local-mac-mini openclaw-security-hardening rag-infrastructure vllm-server ai-inference-service-mesh cdn-setup dns-management llm-gateway load-balancing reverse-proxy service-mesh convex-backend firebase-app-platform vercel-deployments gpu-server-management linux-administration performance-tuning ssh-configuration systemd-services user-management windows-server backup-recovery block-storage nfs-storage object-storage ai-agent-security ai-coding-agent-guardrails ai-red-teaming ai-security-hardening llm-app-security mcp-server-security model-supply-chain-security prompt-injection-defense cis-benchmarks container-hardening kubernetes-hardening linux-hardening openclaw-deployment-hardening windows-hardening firewall-config ssl-tls-management vpn-setup waf-setup zero-trust incident-response penetration-testing security-automation threat-modeling container-scanning dast-scanning dependency-scanning sast-scanning sbom-supply-chain supply-chain-attack-response vulnerability-scanning aws-secrets-manager azure-keyvault gcp-secret-manager hashicorp-vault sops-encryption -a claude-code -y

Утилита skills ставит текущую версию из репозитория. Чтобы скилл работал во всех проектах, добавьте флаг -g.

Выполните в терминале в папке проекта

npx skills add bagelhole/devops-security-agent-skills --skill audit-logging aws-cloudtrail azure-monitor-audit gcp-audit-logs business-continuity disaster-recovery incident-management runbook-creation fedramp-compliance gdpr-compliance hipaa-compliance iso27001-compliance pci-dss-compliance soc2-compliance access-review asset-inventory change-management policy-as-code vendor-management agent-evals agent-observability ai-pipeline-orchestration ai-sre-incident-response llm-caching llm-cost-optimization llmops-platform-engineering model-registry-governance rag-observability-evals azure-devops circleci github-actions gitlab-ci jenkins container-registries docker-compose docker-management podman devcontainers-nix alerting-oncall datadog ebpf-observability elk-stack loki-logging new-relic opentelemetry prometheus-grafana sre-dashboards argocd-gitops helm-charts kubernetes-ops kustomize model-serving-kubernetes openshift platform-engineering blue-green-deploy feature-flags git-workflow semantic-versioning aws-cost-optimization aws-ec2 aws-ecs-fargate aws-iam aws-lambda aws-rds aws-s3 aws-vpc cloudformation terraform-aws arm-templates azure-aks azure-functions azure-networking azure-sql azure-vms terraform-azure gcp-cloud-functions gcp-cloud-sql gcp-compute gcp-gke gcp-networking terraform-gcp cloudflare-pages cloudflare-r2 cloudflare-workers cloudflare-zero-trust database-backups mongodb mysql planetscale postgresql redis vector-database-ops opentofu-migration identity-access-management mdm-device-management saas-security-posture startup-it-troubleshooting gpu-kubernetes-operations llm-fine-tuning llm-inference-scaling mac-mini-llm-lab multi-tenant-llm-hosting ollama-stack openclaw-local-mac-mini openclaw-security-hardening rag-infrastructure vllm-server ai-inference-service-mesh cdn-setup dns-management llm-gateway load-balancing reverse-proxy service-mesh convex-backend firebase-app-platform vercel-deployments gpu-server-management linux-administration performance-tuning ssh-configuration systemd-services user-management windows-server backup-recovery block-storage nfs-storage object-storage ai-agent-security ai-coding-agent-guardrails ai-red-teaming ai-security-hardening llm-app-security mcp-server-security model-supply-chain-security prompt-injection-defense cis-benchmarks container-hardening kubernetes-hardening linux-hardening openclaw-deployment-hardening windows-hardening firewall-config ssl-tls-management vpn-setup waf-setup zero-trust incident-response penetration-testing security-automation threat-modeling container-scanning dast-scanning dependency-scanning sast-scanning sbom-supply-chain supply-chain-attack-response vulnerability-scanning aws-secrets-manager azure-keyvault gcp-secret-manager hashicorp-vault sops-encryption -a cursor -y

Утилита skills ставит текущую версию из репозитория. Чтобы скилл работал во всех проектах, добавьте флаг -g.

Выполните в терминале в папке проекта

npx skills add bagelhole/devops-security-agent-skills --skill audit-logging aws-cloudtrail azure-monitor-audit gcp-audit-logs business-continuity disaster-recovery incident-management runbook-creation fedramp-compliance gdpr-compliance hipaa-compliance iso27001-compliance pci-dss-compliance soc2-compliance access-review asset-inventory change-management policy-as-code vendor-management agent-evals agent-observability ai-pipeline-orchestration ai-sre-incident-response llm-caching llm-cost-optimization llmops-platform-engineering model-registry-governance rag-observability-evals azure-devops circleci github-actions gitlab-ci jenkins container-registries docker-compose docker-management podman devcontainers-nix alerting-oncall datadog ebpf-observability elk-stack loki-logging new-relic opentelemetry prometheus-grafana sre-dashboards argocd-gitops helm-charts kubernetes-ops kustomize model-serving-kubernetes openshift platform-engineering blue-green-deploy feature-flags git-workflow semantic-versioning aws-cost-optimization aws-ec2 aws-ecs-fargate aws-iam aws-lambda aws-rds aws-s3 aws-vpc cloudformation terraform-aws arm-templates azure-aks azure-functions azure-networking azure-sql azure-vms terraform-azure gcp-cloud-functions gcp-cloud-sql gcp-compute gcp-gke gcp-networking terraform-gcp cloudflare-pages cloudflare-r2 cloudflare-workers cloudflare-zero-trust database-backups mongodb mysql planetscale postgresql redis vector-database-ops opentofu-migration identity-access-management mdm-device-management saas-security-posture startup-it-troubleshooting gpu-kubernetes-operations llm-fine-tuning llm-inference-scaling mac-mini-llm-lab multi-tenant-llm-hosting ollama-stack openclaw-local-mac-mini openclaw-security-hardening rag-infrastructure vllm-server ai-inference-service-mesh cdn-setup dns-management llm-gateway load-balancing reverse-proxy service-mesh convex-backend firebase-app-platform vercel-deployments gpu-server-management linux-administration performance-tuning ssh-configuration systemd-services user-management windows-server backup-recovery block-storage nfs-storage object-storage ai-agent-security ai-coding-agent-guardrails ai-red-teaming ai-security-hardening llm-app-security mcp-server-security model-supply-chain-security prompt-injection-defense cis-benchmarks container-hardening kubernetes-hardening linux-hardening openclaw-deployment-hardening windows-hardening firewall-config ssl-tls-management vpn-setup waf-setup zero-trust incident-response penetration-testing security-automation threat-modeling container-scanning dast-scanning dependency-scanning sast-scanning sbom-supply-chain supply-chain-attack-response vulnerability-scanning aws-secrets-manager azure-keyvault gcp-secret-manager hashicorp-vault sops-encryption -a github-copilot -y

Утилита skills ставит текущую версию из репозитория. Чтобы скилл работал во всех проектах, добавьте флаг -g.

Выполните в терминале в папке проекта

npx skills add bagelhole/devops-security-agent-skills --skill audit-logging aws-cloudtrail azure-monitor-audit gcp-audit-logs business-continuity disaster-recovery incident-management runbook-creation fedramp-compliance gdpr-compliance hipaa-compliance iso27001-compliance pci-dss-compliance soc2-compliance access-review asset-inventory change-management policy-as-code vendor-management agent-evals agent-observability ai-pipeline-orchestration ai-sre-incident-response llm-caching llm-cost-optimization llmops-platform-engineering model-registry-governance rag-observability-evals azure-devops circleci github-actions gitlab-ci jenkins container-registries docker-compose docker-management podman devcontainers-nix alerting-oncall datadog ebpf-observability elk-stack loki-logging new-relic opentelemetry prometheus-grafana sre-dashboards argocd-gitops helm-charts kubernetes-ops kustomize model-serving-kubernetes openshift platform-engineering blue-green-deploy feature-flags git-workflow semantic-versioning aws-cost-optimization aws-ec2 aws-ecs-fargate aws-iam aws-lambda aws-rds aws-s3 aws-vpc cloudformation terraform-aws arm-templates azure-aks azure-functions azure-networking azure-sql azure-vms terraform-azure gcp-cloud-functions gcp-cloud-sql gcp-compute gcp-gke gcp-networking terraform-gcp cloudflare-pages cloudflare-r2 cloudflare-workers cloudflare-zero-trust database-backups mongodb mysql planetscale postgresql redis vector-database-ops opentofu-migration identity-access-management mdm-device-management saas-security-posture startup-it-troubleshooting gpu-kubernetes-operations llm-fine-tuning llm-inference-scaling mac-mini-llm-lab multi-tenant-llm-hosting ollama-stack openclaw-local-mac-mini openclaw-security-hardening rag-infrastructure vllm-server ai-inference-service-mesh cdn-setup dns-management llm-gateway load-balancing reverse-proxy service-mesh convex-backend firebase-app-platform vercel-deployments gpu-server-management linux-administration performance-tuning ssh-configuration systemd-services user-management windows-server backup-recovery block-storage nfs-storage object-storage ai-agent-security ai-coding-agent-guardrails ai-red-teaming ai-security-hardening llm-app-security mcp-server-security model-supply-chain-security prompt-injection-defense cis-benchmarks container-hardening kubernetes-hardening linux-hardening openclaw-deployment-hardening windows-hardening firewall-config ssl-tls-management vpn-setup waf-setup zero-trust incident-response penetration-testing security-automation threat-modeling container-scanning dast-scanning dependency-scanning sast-scanning sbom-supply-chain supply-chain-attack-response vulnerability-scanning aws-secrets-manager azure-keyvault gcp-secret-manager hashicorp-vault sops-encryption -a codex -y

Утилита skills ставит текущую версию из репозитория. Чтобы скилл работал во всех проектах, добавьте флаг -g.

Выполните в терминале в папке проекта

npx skills add bagelhole/devops-security-agent-skills --skill audit-logging aws-cloudtrail azure-monitor-audit gcp-audit-logs business-continuity disaster-recovery incident-management runbook-creation fedramp-compliance gdpr-compliance hipaa-compliance iso27001-compliance pci-dss-compliance soc2-compliance access-review asset-inventory change-management policy-as-code vendor-management agent-evals agent-observability ai-pipeline-orchestration ai-sre-incident-response llm-caching llm-cost-optimization llmops-platform-engineering model-registry-governance rag-observability-evals azure-devops circleci github-actions gitlab-ci jenkins container-registries docker-compose docker-management podman devcontainers-nix alerting-oncall datadog ebpf-observability elk-stack loki-logging new-relic opentelemetry prometheus-grafana sre-dashboards argocd-gitops helm-charts kubernetes-ops kustomize model-serving-kubernetes openshift platform-engineering blue-green-deploy feature-flags git-workflow semantic-versioning aws-cost-optimization aws-ec2 aws-ecs-fargate aws-iam aws-lambda aws-rds aws-s3 aws-vpc cloudformation terraform-aws arm-templates azure-aks azure-functions azure-networking azure-sql azure-vms terraform-azure gcp-cloud-functions gcp-cloud-sql gcp-compute gcp-gke gcp-networking terraform-gcp cloudflare-pages cloudflare-r2 cloudflare-workers cloudflare-zero-trust database-backups mongodb mysql planetscale postgresql redis vector-database-ops opentofu-migration identity-access-management mdm-device-management saas-security-posture startup-it-troubleshooting gpu-kubernetes-operations llm-fine-tuning llm-inference-scaling mac-mini-llm-lab multi-tenant-llm-hosting ollama-stack openclaw-local-mac-mini openclaw-security-hardening rag-infrastructure vllm-server ai-inference-service-mesh cdn-setup dns-management llm-gateway load-balancing reverse-proxy service-mesh convex-backend firebase-app-platform vercel-deployments gpu-server-management linux-administration performance-tuning ssh-configuration systemd-services user-management windows-server backup-recovery block-storage nfs-storage object-storage ai-agent-security ai-coding-agent-guardrails ai-red-teaming ai-security-hardening llm-app-security mcp-server-security model-supply-chain-security prompt-injection-defense cis-benchmarks container-hardening kubernetes-hardening linux-hardening openclaw-deployment-hardening windows-hardening firewall-config ssl-tls-management vpn-setup waf-setup zero-trust incident-response penetration-testing security-automation threat-modeling container-scanning dast-scanning dependency-scanning sast-scanning sbom-supply-chain supply-chain-attack-response vulnerability-scanning aws-secrets-manager azure-keyvault gcp-secret-manager hashicorp-vault sops-encryption -a gemini-cli -y

Утилита skills ставит текущую версию из репозитория. Чтобы скилл работал во всех проектах, добавьте флаг -g.

Выполните в терминале в папке проекта

npx skills add bagelhole/devops-security-agent-skills --skill audit-logging aws-cloudtrail azure-monitor-audit gcp-audit-logs business-continuity disaster-recovery incident-management runbook-creation fedramp-compliance gdpr-compliance hipaa-compliance iso27001-compliance pci-dss-compliance soc2-compliance access-review asset-inventory change-management policy-as-code vendor-management agent-evals agent-observability ai-pipeline-orchestration ai-sre-incident-response llm-caching llm-cost-optimization llmops-platform-engineering model-registry-governance rag-observability-evals azure-devops circleci github-actions gitlab-ci jenkins container-registries docker-compose docker-management podman devcontainers-nix alerting-oncall datadog ebpf-observability elk-stack loki-logging new-relic opentelemetry prometheus-grafana sre-dashboards argocd-gitops helm-charts kubernetes-ops kustomize model-serving-kubernetes openshift platform-engineering blue-green-deploy feature-flags git-workflow semantic-versioning aws-cost-optimization aws-ec2 aws-ecs-fargate aws-iam aws-lambda aws-rds aws-s3 aws-vpc cloudformation terraform-aws arm-templates azure-aks azure-functions azure-networking azure-sql azure-vms terraform-azure gcp-cloud-functions gcp-cloud-sql gcp-compute gcp-gke gcp-networking terraform-gcp cloudflare-pages cloudflare-r2 cloudflare-workers cloudflare-zero-trust database-backups mongodb mysql planetscale postgresql redis vector-database-ops opentofu-migration identity-access-management mdm-device-management saas-security-posture startup-it-troubleshooting gpu-kubernetes-operations llm-fine-tuning llm-inference-scaling mac-mini-llm-lab multi-tenant-llm-hosting ollama-stack openclaw-local-mac-mini openclaw-security-hardening rag-infrastructure vllm-server ai-inference-service-mesh cdn-setup dns-management llm-gateway load-balancing reverse-proxy service-mesh convex-backend firebase-app-platform vercel-deployments gpu-server-management linux-administration performance-tuning ssh-configuration systemd-services user-management windows-server backup-recovery block-storage nfs-storage object-storage ai-agent-security ai-coding-agent-guardrails ai-red-teaming ai-security-hardening llm-app-security mcp-server-security model-supply-chain-security prompt-injection-defense cis-benchmarks container-hardening kubernetes-hardening linux-hardening openclaw-deployment-hardening windows-hardening firewall-config ssl-tls-management vpn-setup waf-setup zero-trust incident-response penetration-testing security-automation threat-modeling container-scanning dast-scanning dependency-scanning sast-scanning sbom-supply-chain supply-chain-attack-response vulnerability-scanning aws-secrets-manager azure-keyvault gcp-secret-manager hashicorp-vault sops-encryption -a cline -y

Утилита skills ставит текущую версию из репозитория. Чтобы скилл работал во всех проектах, добавьте флаг -g.

Выполните в терминале в папке проекта

npx skills add bagelhole/devops-security-agent-skills --skill audit-logging aws-cloudtrail azure-monitor-audit gcp-audit-logs business-continuity disaster-recovery incident-management runbook-creation fedramp-compliance gdpr-compliance hipaa-compliance iso27001-compliance pci-dss-compliance soc2-compliance access-review asset-inventory change-management policy-as-code vendor-management agent-evals agent-observability ai-pipeline-orchestration ai-sre-incident-response llm-caching llm-cost-optimization llmops-platform-engineering model-registry-governance rag-observability-evals azure-devops circleci github-actions gitlab-ci jenkins container-registries docker-compose docker-management podman devcontainers-nix alerting-oncall datadog ebpf-observability elk-stack loki-logging new-relic opentelemetry prometheus-grafana sre-dashboards argocd-gitops helm-charts kubernetes-ops kustomize model-serving-kubernetes openshift platform-engineering blue-green-deploy feature-flags git-workflow semantic-versioning aws-cost-optimization aws-ec2 aws-ecs-fargate aws-iam aws-lambda aws-rds aws-s3 aws-vpc cloudformation terraform-aws arm-templates azure-aks azure-functions azure-networking azure-sql azure-vms terraform-azure gcp-cloud-functions gcp-cloud-sql gcp-compute gcp-gke gcp-networking terraform-gcp cloudflare-pages cloudflare-r2 cloudflare-workers cloudflare-zero-trust database-backups mongodb mysql planetscale postgresql redis vector-database-ops opentofu-migration identity-access-management mdm-device-management saas-security-posture startup-it-troubleshooting gpu-kubernetes-operations llm-fine-tuning llm-inference-scaling mac-mini-llm-lab multi-tenant-llm-hosting ollama-stack openclaw-local-mac-mini openclaw-security-hardening rag-infrastructure vllm-server ai-inference-service-mesh cdn-setup dns-management llm-gateway load-balancing reverse-proxy service-mesh convex-backend firebase-app-platform vercel-deployments gpu-server-management linux-administration performance-tuning ssh-configuration systemd-services user-management windows-server backup-recovery block-storage nfs-storage object-storage ai-agent-security ai-coding-agent-guardrails ai-red-teaming ai-security-hardening llm-app-security mcp-server-security model-supply-chain-security prompt-injection-defense cis-benchmarks container-hardening kubernetes-hardening linux-hardening openclaw-deployment-hardening windows-hardening firewall-config ssl-tls-management vpn-setup waf-setup zero-trust incident-response penetration-testing security-automation threat-modeling container-scanning dast-scanning dependency-scanning sast-scanning sbom-supply-chain supply-chain-attack-response vulnerability-scanning aws-secrets-manager azure-keyvault gcp-secret-manager hashicorp-vault sops-encryption -a roo -y

Утилита skills ставит текущую версию из репозитория. Чтобы скилл работал во всех проектах, добавьте флаг -g.

Форк Roo Code, папки .roo те же.

Выполните в терминале в папке проекта

npx skills add bagelhole/devops-security-agent-skills --skill audit-logging aws-cloudtrail azure-monitor-audit gcp-audit-logs business-continuity disaster-recovery incident-management runbook-creation fedramp-compliance gdpr-compliance hipaa-compliance iso27001-compliance pci-dss-compliance soc2-compliance access-review asset-inventory change-management policy-as-code vendor-management agent-evals agent-observability ai-pipeline-orchestration ai-sre-incident-response llm-caching llm-cost-optimization llmops-platform-engineering model-registry-governance rag-observability-evals azure-devops circleci github-actions gitlab-ci jenkins container-registries docker-compose docker-management podman devcontainers-nix alerting-oncall datadog ebpf-observability elk-stack loki-logging new-relic opentelemetry prometheus-grafana sre-dashboards argocd-gitops helm-charts kubernetes-ops kustomize model-serving-kubernetes openshift platform-engineering blue-green-deploy feature-flags git-workflow semantic-versioning aws-cost-optimization aws-ec2 aws-ecs-fargate aws-iam aws-lambda aws-rds aws-s3 aws-vpc cloudformation terraform-aws arm-templates azure-aks azure-functions azure-networking azure-sql azure-vms terraform-azure gcp-cloud-functions gcp-cloud-sql gcp-compute gcp-gke gcp-networking terraform-gcp cloudflare-pages cloudflare-r2 cloudflare-workers cloudflare-zero-trust database-backups mongodb mysql planetscale postgresql redis vector-database-ops opentofu-migration identity-access-management mdm-device-management saas-security-posture startup-it-troubleshooting gpu-kubernetes-operations llm-fine-tuning llm-inference-scaling mac-mini-llm-lab multi-tenant-llm-hosting ollama-stack openclaw-local-mac-mini openclaw-security-hardening rag-infrastructure vllm-server ai-inference-service-mesh cdn-setup dns-management llm-gateway load-balancing reverse-proxy service-mesh convex-backend firebase-app-platform vercel-deployments gpu-server-management linux-administration performance-tuning ssh-configuration systemd-services user-management windows-server backup-recovery block-storage nfs-storage object-storage ai-agent-security ai-coding-agent-guardrails ai-red-teaming ai-security-hardening llm-app-security mcp-server-security model-supply-chain-security prompt-injection-defense cis-benchmarks container-hardening kubernetes-hardening linux-hardening openclaw-deployment-hardening windows-hardening firewall-config ssl-tls-management vpn-setup waf-setup zero-trust incident-response penetration-testing security-automation threat-modeling container-scanning dast-scanning dependency-scanning sast-scanning sbom-supply-chain supply-chain-attack-response vulnerability-scanning aws-secrets-manager azure-keyvault gcp-secret-manager hashicorp-vault sops-encryption -a opencode -y

Утилита skills ставит текущую версию из репозитория. Чтобы скилл работал во всех проектах, добавьте флаг -g.

Понадобится: Node.js

Сверено с репозиторием 26 сент. 2026, коммит 0365f57.

Текст для агента

Установи набор командой npx skills add bagelhole/DevOps-Security-Agent-Skills, выбери нужные скиллы. Можно клонировать репозиторий в каталог скиллов вручную.

Другие способы из описания автора
npx skills add bagelhole/DevOps-Security-Agent-Skills

Ставит скиллы в каталог агента; отдельные скиллы через --skill.

Это чужой код. Посмотрите файлы в репозитории перед установкой.

Что делает

Набор дает агенту прикладные знания по DevOps и безопасности: Kubernetes, Terraform, облака AWS, Azure и GCP, работа с секретами, укрепление контейнеров, аудит и непрерывность бизнеса. Скиллы сгруппированы по разделам compliance, devops, infrastructure и security, каждый описан в SKILL.md и подходит любому агенту, который читает файлы. Есть отдельные скиллы по безопасности AI: гардрейлы для кодовых агентов, укрепление MCP-серверов, red teaming. Ставится через skills CLI сразу для нескольких агентов или клонированием.

Для кого. Для инженеров DevOps и специалистов по безопасности, работающих через AI-агента.

Подходит, если

  • Нужны знания по Kubernetes, Terraform или облакам в агенте
  • Нужно укрепить контейнеры, секреты или MCP-сервер
  • Нужны материалы по аудиту и комплаенсу

Не подходит, если

  • Нужен запускаемый инструмент, а не справочные знания
  • Работа далека от инфраструктуры и безопасности

Пример запроса к агенту

Проверь мой Terraform и подскажи, как укрепить хранение секретов

Ограничения

Скиллы дают знания и практики, но не выполняют операции за вас и не заменяют проверенные пайплайны и политики. Материал на английском. Полный набор большой, лучше ставить нужные скиллы.

Как отключить. Удалите установленные скиллы через skills CLI или папку с ними из каталога скиллов агента.

Проверка безопасности

  • Скиллы состоят из справочников и практик, операции не выполняют

Коротко о README

README описывает базу из более чем ста скиллов по DevOps, безопасности, инфраструктуре и комплаенсу для любых агентов, читающих файлы. Установка идет через skills CLI сразу для нескольких агентов, отдельные скиллы ставят флагом skill, есть и вариант клонирования. Скиллы разложены по разделам compliance, devops, infrastructure и security, включая безопасность AI и MCP. Лицензия MIT.

SKILL.md

---
name: mcp-server-security
description: Secure Model Context Protocol (MCP) servers with transport encryption, tool authorization, input validation, and audit logging for safe AI agent integrations.
license: MIT
metadata:
  author: devops-skills
  version: "1.0"
---

# MCP Server Security

Comprehensive hardening guide for Model Context Protocol (MCP) servers.

Частые вопросы

С какими агентами работает?

С Claude Code, Cursor, Codex, OpenCode, Cline и другими агентами, которые читают файлы скиллов.

Можно поставить только один скилл?

Да, через флаг --skill в skills CLI, например kubernetes-ops.

Выбор редакции

MCP-сервер, встроенный в агент Netdata: метрики, логи, алерты и живые данные о процессах, сервисах и контейнерах для AI-ассистента

MCP-серверСредний рискБез VPN80,7 тыс.Звезды репозитория
Выбор редакции

Официальный MCP-сервер GitHub: код, issues, пул-реквесты, Actions и алерты безопасности прямо из агента

MCP-серверВысокий риск33,3 тыс.Звезды репозитория

Agent Skills для продуктов Google

Agent Skills for Google products and technologies

Выбор редакции

Официальная коллекция скиллов Google для работы с Google Cloud, BigQuery, GKE, рекламой и аналитикой из агента

СкиллВысокий риск20,5 тыс.Звезды репозитория

MCP-серверы для AWS

Open source MCP servers for AWS

Официальный

Официальный набор MCP-серверов AWS: документация, IaC, контейнеры, serverless, базы данных, стоимость и мониторинг

MCP-серверВысокий риск9,7 тыс.Звезды репозитория
Foxx AIDevOps-Security-Agent-Skills

Я Foxx AI и уже разобрал этот инструмент. Спросите про установку, настройку или что угодно еще, отвечу простыми словами.