azure-specialist
Deep-dive Azure architecture review, debugging, and service design. Use for structured investigations of Azure-specific issues, identity or cost audits, and multi-service design reviews. Triggers on: "Azure audit", "Azure design review", "EntraID review", "Managed Identity debug", "review my Azure", "Azure troubleshooting", "AKS deep-dive".
- License 许可
- MIT
- 最近更新
- 2026-10-26
- Stars
- ★ 0
- Forks
- 0
01 前言 Front Matter
- id
- azure-specialist
- name
- azure-specialist
- task
- [ "debugging", "review" ]
- type
- skill
- model
- opus
- owner
- @kaiohenricunha
- tools
- Read, Grep, Glob, Bash
- domain
- [ "infra" ]
- effort
- max
- created
- 1735689600000
- updated
- 1776384000000
- version
- 1.0.0
- dotbabel
- { "compute": { "mode": "dynamic", "binding": "consumer", "rationale": "Evidence-based investigation or evaluation must hold several technical constraints at once.", "requirement": "deep" } }
- maturity
- validated
- platform
- [ "azure" ]
- description
- Deep-dive Azure architecture review, debugging, and service design. Use for structured investigations of Azure-specific issues, identity or cost audits, and multi-service design reviews. Triggers on: "Azure audit", "Azure design review", "EntraID review", "Managed Identity debug", "review my Azure", "Azure troubleshooting", "AKS deep-dive".
- allowed-tools
- Read Grep Glob Bash
- argument-hint
- <subscription context, service, or problem description>
02 资源文件 7
- 参考文档references/compute.md2.8 KB
- 参考文档references/devops.md2.9 KB
- 参考文档references/iac-patterns.md2.9 KB
- 参考文档references/identity.md3.2 KB
- 参考文档references/networking.md3.0 KB
- 参考文档references/serverless.md2.8 KB
- 参考文档references/storage.md3.3 KB
03 SKILL.md 正文
Azure Specialist
Structured investigation for Azure workloads. Five phases: gather context, diagnose, design, recommend, verify.
Arguments
$0— subscription context, service scope, or problem description. Required.
Phase 1: Context Gathering
- Identify the tenant, subscription(s), resource group(s), and services in scope.
- Glob for IaC in the working directory:
**/*.bicep,**/azuredeploy.json,**/main.tf,**/*.parameters.json. - If Azure CLI access is available:
az account show az group list --query "[].name" - Note which resource providers are registered:
az provider list --query "[?registrationState=='Registered'].namespace" -o tsv
Phase 2: Diagnosis
Compute / containers:
az vm list --query "[].{name:name,rg:resourceGroup,state:powerState}" -o table
az aks list --query "[].{name:name,rg:resourceGroup,version:kubernetesVersion}" -o table
az webapp list --query "[].{name:name,rg:resourceGroup,state:state}" -o table
Identity / RBAC:
az role assignment list --assignee <principal-id> --all
az ad app list --display-name <name>
az identity list --resource-group <rg>
Networking:
az network vnet list
az network nsg rule list --nsg-name <nsg> --resource-group <rg>
az network private-endpoint list
Serverless / events:
az functionapp list
az servicebus namespace list
az eventgrid topic list
Cost / quotas:
az consumption usage list --start-date <> --end-date <>
az vm list-usage --location <region>
Phase 3: Design / Root-Cause Analysis
Map symptoms to causes:
| Symptom | Common Causes | Check |
|---|---|---|
| AKS pod AuthN fails | Managed Identity not assigned, missing federated credential | az aks show --query identity + pod ServiceAccount annotations |
| App Service slow | Cold start on consumption plan, misconfigured scale rules | Plan tier, autoscale settings |
| Storage 403 | Private endpoint with wrong DNS, firewall IP allowlist | az storage account network-rule list |
| Function cold starts | Consumption plan + infrequent traffic | Switch to Premium or Always-Ready instances |
| Cosmos DB throttle (429) | RU/s too low, hot partition | Diagnostic settings, metrics, partition key review |
| EntraID app login fails | Redirect URI mismatch, missing API permission grant | az ad app show + consent status |
Cite resource ID or file:line for every finding.
Phase 4: Recommendations
Output findings in priority order:
[CRITICAL] <title>
Resource: <resource ID or file:line>
Issue: <one sentence>
Evidence: <CLI output or code snippet>
Fix: <specific change, with Bicep/ARM/Terraform diff>
Trade-off: <alternative and its downside, if meaningful>
- Order: CRITICAL → WARNING → INFO.
- For IaC fixes, show the exact Bicep/ARM/Terraform diff.
- Reference relevant docs in
references/where applicable.
Phase 5: Verification
After fixes are applied:
- Re-run the diagnostic command that surfaced the issue.
- For RBAC/Managed Identity changes: verify with
az role assignment listand a live workload token request. - For network changes:
az network watcher test-connectivityor NSG flow-log review. - For Bicep/ARM deployments: run
what-ifbefore and after to confirm intended drift only. - Check Azure Monitor metrics and Service Health — no new alerts should be firing.
Reference Docs
Consult references/ for decision guides:
| File | When to use |
|---|---|
compute.md |
AKS, ACI, VMs, App Service |
serverless.md |
Functions, Logic Apps, Service Bus, Event Grid |
storage.md |
Blob, Files, Queues, Cosmos DB, Azure SQL |
networking.md |
VNet, App Gateway, Front Door, Private Endpoints |
identity.md |
EntraID, Managed Identity, RBAC scopes |
devops.md |
Azure Pipelines, ACR, release management |
iac-patterns.md |
Bicep, ARM, Terraform (AzureRM/azapi) patterns |
04 同仓库更多 Skills
yolo
Start temporary autonomy for the current task or an explicit scope with a required --for expiry. Codex permits scoped pushes, merges, deploys, and releases after strict testing and validation; Full access is a separate runtime setting. Claude Code uses a session permission hook with a protected-action denylist.
veracity-audit
Audit a data pipeline for Veracity and Value. Dispatches data-scientist, compliance-auditor, and data-engineer agents with project context injected at dispatch time. All source paths are supplied via flags — no defaults, no project assumptions. Subcommands: audit (full pipeline walk), score-check (scoring math only), gate-check (gate coverage only), source-trace <label> (single source end-to-end). Invoke when: "audit pipeline", "veracity check", "scoring math correct?", "check gates", "trace source", "verify quality gates", "formula correct?".
validate-spec
Audit an already-implemented spec against the codebase. Walks each constraint (ARCH-N, PERF-N, KD-N, etc.) and acceptance criterion, grounds findings in file:line evidence, runs the spec.json acceptance_commands, and writes a single audit doc to docs/audits/. Use whenever the user asks to "validate a spec", "audit a spec", "check if spec is implemented", "verify the spec is done", "is this spec really done", or otherwise wants closure on spec-driven work. Read-only against the spec — produces an audit, never modifies the spec itself.
terragrunt-specialist
Deep-dive Terragrunt hierarchy review, DRY pattern audits, and run-all orchestration analysis. Use for structured investigations of multi-environment Terragrunt layouts, dependency graphs, remote state config, and hook correctness. Triggers on: "Terragrunt audit", "run-all review", "dependency block", "DRY pattern review", "env hierarchy audit", "mock_outputs", "terragrunt hooks".
terraform-specialist
Deep-dive Terraform architecture review, module design, state management, and migration. Use for structured investigations of Terraform workspaces, provider configuration, module coupling, import workflows, and test coverage. Triggers on: "Terraform audit", "module review", "state management", "Terraform import", "workspace design", "provider config review", "Terraform testing".
spec
Create structured engineering specs through interactive pairing. Use when the user wants to create a spec, design doc, technical specification, architecture document, RFC, or engineering plan. Triggers on "let's spec this out", "create a spec", "design doc", "write a technical plan", "plan the architecture". Works for greenfield and brownfield projects. Outputs to docs/specs/.