Azure Build

10 skills inclus

• 01

  Azure Prepare

  Fait préparer par l'agent un projet pour Azure avec azd : plan de déploiement validé d'abord, puis azure.yaml, Bicep ou Terraform, Dockerfiles et sécurité.+

  sf-azure-prepare/SKILL.md, de microsoft/azure-skills, licence MIT.

  1---

  2name: sf-azure-prepare

  3description: "Prepares a project for deployment to Azure with the Azure Developer CLI (azd): writes a deploym…

  4---

  5

  6# Azure Prepare

  7

  8Turn an application into a deployable Azure project: plan first, get approval, then generate infrastructure a…

  9

  10## When to use

  11

  12- Create a new app for Azure, add services to an existing one, modernize or migrate an app to Azure.

  13- Set up Azure infrastructure (azd, Azure CLI, Bicep or Terraform), including "create and deploy" requests: p…

  14

  15Not for: Python App Service code-only deploys (`sf-azure-python-appservice`), running `azd up` or `terraform…

  16
• 02

  Azure Validate

  Fait valider par l'agent un projet Azure avant déploiement via un script pas à pas : config, Bicep ou Terraform, rôles RBAC, identités managées.+

  sf-azure-validate/SKILL.md, de microsoft/azure-skills, licence MIT.

  1---

  2name: sf-azure-validate

  3description: "Validates an Azure project before deployment: configuration, azure.yaml, Bicep or Terraform (in…

  4---

  5

  6# Azure Validate

  7

  8Prove that a prepared project can deploy before anything is deployed. This skill is the only step allowed to…

  9

  10## Prerequisite

  11

  12`sf-azure-prepare` must have run: `.azure/deployment-plan.md` exists with status `Approved` or later. If the…

  13

  14## When to use

  15

  16- "Is my app ready to deploy?", preflight checks, validate `azure.yaml`, Bicep or Terraform, what-if analysis.
• 03

  Azure Deploy

  Fait exécuter par l'agent le déploiement Azure d'un projet préparé et validé : checklist, azd up ou terraform apply, reprise d'erreur, vérification des rôles.+

  sf-azure-deploy/SKILL.md, de microsoft/azure-skills, licence MIT.

  1---

  2name: sf-azure-deploy

  3description: "Executes Azure deployments for already prepared and validated projects (existing .azure/deploym…

  4---

  5

  6# Azure Deploy

  7

  8Run the deployment of a project that `sf-azure-prepare` prepared and `sf-azure-validate` validated, recover f…

  9

  10## Prerequisites

  11

  12Both must hold, or stop:

  13

  141. `.azure/deployment-plan.md` exists (from `sf-azure-prepare`). No plan: run `sf-azure-prepare`.

  152. Its status is `Validated` and its Validation Proof section holds real results with commands and timestamps…

  16
• 04

  Azure App Readiness

  Fait évaluer par l'agent, en lecture seule, si le code d'un dépôt est prêt pour Azure : build, complétude, dépendances, frameworks, bloquants.+

  sf-azure-app-readiness/SKILL.md, de microsoft/azure-skills, licence MIT.

  1---

  2name: sf-azure-app-readiness

  3description: "Assesses whether a repository's source code is ready to deploy to Azure before any infrastructu…

  4---

  5

  6# Azure App Readiness

  7

  8Evaluate a repository for build health, completeness and Azure deployment feasibility, before infrastructure…

  9

  10Upstream, this skill is phase 1 of a four-phase onboarding pipeline run by an `azure-app-onboard` orchestrato…

  11

  12## When not to use

  13

  14| Need | Use |

  15|---|---|

  16| Validate infrastructure (Bicep, Terraform, `azure.yaml`) | `sf-azure-validate` |
• 05

  Python on Azure App Service

  Fait déployer par l'agent une app Python (Flask, Django, FastAPI) sur App Service Linux, ressources créées si besoin après accord, URL rendue.+

  sf-azure-python-appservice/SKILL.md, de microsoft/azure-skills, licence MIT.

  1---

  2name: sf-azure-python-appservice

  3description: "Deploys Python web app code (Flask, Django, FastAPI or generic) to Azure App Service on Linux w…

  4---

  5

  6# Python on Azure App Service

  7

  8Deploy Python code to Azure App Service Linux (plan P0v3, Python 3.14), creating the resource group, plan and…

  9

  10When the Azure MCP server is available, use `mcp_azure_mcp_subscription_list`, `mcp_azure_mcp_group_list`, `m…

  11

  12## Process

  13

  141. **Resolve context with defaults.** Only the app name is asked; derive resource group `<app>-rg`, plan `<ap…

  152. **Detect the framework** (advisory, never blocking): [detect.md](references/detect.md).

  163. **Choose the path.** `azure.yaml` with an App Service host: [deploy-azd.md](references/deploy-azd.md); oth…
• 06

  Azure VMs

  Fait orienter par l'agent les demandes de VM et VMSS vers le bon workflow : recommandation et prix, création, réservation de capacité, enrôlement.+

  sf-azure-vms/SKILL.md, de microsoft/azure-skills, licence MIT.

  1---

  2name: sf-azure-vms

  3description: "Routes Azure virtual machine and scale set (VMSS) requests to the right workflow: recommending…

  4---

  5

  6# Azure VMs

  7

  8Route VM and VMSS requests to one workflow, then load only the references that workflow asks for.

  9

  10## When to use

  11

  12- Recommend, compare or price a VM or VMSS.

  13- Create, provision or deploy a bare VM or VMSS.

  14- Capacity Reservation Groups (CRG): reserve or guarantee capacity, associate or disassociate.

  15- Essential Machine Management (EMM): machine enrollment and monitoring.

  16
• 07

  Azure Storage

  Fait utiliser par l'agent Azure Storage (Blob, Files, Queues, Tables, Data Lake) via MCP ou CLI et choisir niveaux d'accès et redondance.+

  sf-azure-storage/SKILL.md, de microsoft/azure-skills, licence MIT.

  1---

  2name: sf-azure-storage

  3description: "Works with Azure Storage: Blob Storage, File Shares, Queue Storage, Table Storage and Data Lake…

  4---

  5

  6# Azure Storage

  7

  8Pick the right storage service and tier, then operate it through the Azure MCP server when available, or the…

  9

  10## Services

  11

  12| Service | Use when | MCP | CLI |

  13|---|---|---|---|

  14| Blob Storage | Objects, files, backups, static content | `azure__storage` | `az storage blob` |

  15| File Shares | SMB shares, lift-and-shift | | `az storage file` |

  16| Queue Storage | Async messaging, task queues | | `az storage queue` |
• 08

  Azure Messaging SDK Troubleshooting

  Fait diagnostiquer par l'agent les erreurs des SDK Event Hubs et Service Bus : connexion, AMQP, verrous, sessions, checkpoints, configuration.+

  sf-azure-messaging-debug/SKILL.md, de microsoft/azure-skills, licence MIT.

  1---

  2name: sf-azure-messaging-debug

  3description: "Troubleshoots Azure Event Hubs and Service Bus SDK problems in .NET, Java, JavaScript and Pytho…

  4---

  5

  6# Azure Messaging SDK Troubleshooting

  7

  8Diagnose and fix Event Hubs and Service Bus client problems from the error, the SDK and the namespace's healt…

  9

  10## When to use

  11

  12- Connection failures, authentication errors, AMQP link detach or detach-forced errors.

  13- Idle timeout, connection inactivity, slow reconnection.

  14- Message lock or session lock lost or expired, renewal failures, batch lock timeouts.

  15- An event processor or message handler that stops processing; duplicate events; checkpoint offset resets.

  16- Configuration questions: retry, prefetch, batch size, receive batch behavior, logging.
• 09

  Azure Kubernetes Service

  Fait planifier et configurer par l'agent un cluster AKS de production : choix Day-0, réseau, sécurité, observabilité, mises à jour, coûts, autoscaling.+

  sf-azure-aks/SKILL.md, de microsoft/azure-skills, licence MIT.

  1---

  2name: sf-azure-aks

  3description: "Plans, creates and configures production-ready Azure Kubernetes Service (AKS) clusters: a Day-0…

  4---

  5

  6# Azure Kubernetes Service

  7

  8Produce a recommended AKS cluster configuration from the user's requirements, separating **Day-0 decisions**…

  9

  10## When to use

  11

  12- Create or plan an AKS cluster for production, get a Day-0 checklist.

  13- Design networking (API server access, pod IP model, egress, ingress), identity and secrets, governance (Azu…

  14- Choose between AKS Automatic and Standard.

  15- Optimize: pod rightsizing, VPA, cluster autoscaler, Spot node pools.

  16
• 10

  AI Runway on AKS

  Fait installer par l'agent AI Runway sur un cluster AKS existant jusqu'à un modèle en service : contrôleur, GPU, fournisseur d'inférence, premier déploiement.+

  sf-aks-ai-runway/SKILL.md, de microsoft/azure-skills, licence MIT.

  1---

  2name: sf-aks-ai-runway

  3description: "Sets up AI Runway on an existing AKS cluster, from bare cluster to a running model: cluster ver…

  4argument-hint: "[skip-to-step N]"

  5---

  6

  7# AI Runway on AKS

  8

  9Walk the user from a bare Kubernetes cluster to a running AI model deployment, one step at a time. Resume at…

  10

  11**Cost:** GPU node pools are expensive (an A100 80 GB node can cost several dollars per hour). Make sure the…

  12

  13## Prerequisites

  14

  15An AKS cluster must exist. If not, use `sf-azure-aks` first to create one (with a GPU node pool unless CPU-on…

  16

Extraits. Les dossiers complets sont livrés après le paiement.

Installation