agentleFS
ricmmartinsricmmartins/azure-sre-agent-skills/skills/07-digital-native-governance/SKILL.md

digital-native-governance-check

Skill70 starsChanged 4 months agoMITView on GitHub

Assess Azure governance maturity for Digital Natives and startups. Detects the most common anti-patterns from the "Digital Natives on Azure" series: incomplete Service Health alerts, flat subscription topology without Management Groups, excessive RBAC (everyone is Owner), missing budget alerts, no Managed Identity usage, unprotected web apps without WAF/DDoS, and missing environment separation. Produces a scored maturity report (Foundation → Developing → Established → Optimized). Use when asked about startup readiness, governance maturity, "are we production-ready?", or before a customer security questionnaire.

The file

We split this skill into the rules it gives an agent

6 rules. About 0 are common, 4 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    After first incident: "what basics did we miss?"

    line 27
  2. Distinctive:

    Run all checks in parallel where possible.

    line 41
  3. Very specific:

    Look for prod/nonprod/dev/staging patterns in subscription names.

    line 125
  4. Distinctive:

    If the user provides a list of accepted exceptions, do not flag those items.

    line 364
  5. Distinctive:

    When exceptions are provided:

    line 374
  6. Very specific:

    Use GetAzCliHelp to validate the command syntax before suggesting

    line 418

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Digital Native Governance Check

Purpose

Evaluate whether a startup's Azure environment has the foundational governance that enterprise customers, investors, and auditors expect. Produces a maturity score (0–100) with clear next steps.

This is NOT a deep compliance audit (use the Compliance & Governance skill for that). This is a focused "did you get the basics right?" diagnostic for teams of 5–50 engineers who set up Azure fast and moved on.

When to use this skill

  • Startup preparing for first enterprise customer security review
  • CTO asks "are we production-ready?"
  • Pre-Series A infrastructure maturity check
  • After first incident: "what basics did we miss?"
  • Quarterly governance health check for Digital Natives

Maturity levels

ScoreLevelMeaning
0–39🔴 FoundationGovernance basics missing — will fail any security questionnaire
40–69🟡 DevelopingSome controls exist but significant gaps remain
70–89🟢 EstablishedSolid posture — can pass most security reviews
90–100🏆 OptimizedEnterprise-grade governance on Azure

Assessment procedure

Run all checks in parallel where possible. Score each and sum for total.


🏥 CATEGORY 1 — Alerting & Observability (20 points)

Detects: "Azure VM went down and nobody knew" + "you configured one alert type out of four"

Check 1.1 — Service Health alerts exist (8 pts)

az monitor activity-log alert list \
  --subscription <sub-id> \
  --query "[?contains(to_string(condition.allOf), 'ServiceHealth')].{name:name, enabled:enabled}" \
  -o json

Scoring:

  • 8 pts: Alert exists covering all 4 event types (Service Issues, Health Advisories, Planned Maintenance, Security Advisories)
  • 4 pts: Alert exists but covers only 1–2 types
  • 0 pts: No Service Health alerts at all

📖 https://learn.microsoft.com/en-us/azure/service-health/alerts-activity-log-service-notifications-portal

Check 1.2 — Resource Health alerts for compute (6 pts)

az monitor activity-log alert list \
  --subscription <sub-id> \
  --query "[?contains(to_string(condition.allOf), 'ResourceHealth')].{name:name, enabled:enabled}" \
  -o json

Scoring:

  • 6 pts: Resource Health alerts exist and are enabled
  • 0 pts: No Resource Health alerts — VMs can go Unavailable silently

📖 https://learn.microsoft.com/en-us/azure/service-health/resource-health-alert-monitor-guide

Check 1.3 — Action Groups with valid receivers (6 pts)

az monitor action-group list \
  --subscription <sub-id> \
  --query "[].{name:name, emails:length(emailReceivers), webhooks:length(webhookReceivers)}" \
  -o table

Scoring:

  • 6 pts: Action group with email + at least one other channel (webhook, SMS)
  • 3 pts: Action group with email only
  • 0 pts: No action groups configured

📖 https://learn.microsoft.com/en-us/azure/azure-monitor/alerts/action-groups


🏗️ CATEGORY 2 — Subscription Topology (20 points)

Detects: "the flat subscription problem" + "dev/test/prod in the same subscription"

Check 2.1 — Management Groups beyond Tenant Root (8 pts)

az account management-group list \
  --query "[].{name:name, displayName:displayName}" \
  -o table

Scoring:

  • 8 pts: Custom MG hierarchy exists (nonprod/prod/platform pattern)
  • 4 pts: At least one custom MG beyond Tenant Root Group
  • 0 pts: Only Tenant Root Group — completely flat

📖 https://learn.microsoft.com/en-us/azure/governance/management-groups/overview

Check 2.2 — Environment separation (8 pts)

az account list --all \
  --query "[?state=='Enabled'].{name:name, id:id}" \
  -o table

Look for prod/nonprod/dev/staging patterns in subscription names.

Scoring:

  • 8 pts: Clear separation (prod ≠ nonprod in distinct subscriptions)
  • 4 pts: Multiple subscriptions but unclear separation
  • 0 pts: Single subscription for everything

📖 https://learn.microsoft.com/en-us/azure/cloud-adoption-framework/ready/landing-zone/design-area/resource-org-subscriptions

Check 2.3 — Resource Locks on critical infrastructure (4 pts)

az lock list --subscription <sub-id> \
  --query "[].{name:name, level:level, notes:notes}" \
  -o table

Scoring:

  • 4 pts: CanNotDelete locks on infra resources (Log Analytics, Key Vault, Terraform state)
  • 2 pts: Locks exist but only on non-critical resources
  • 0 pts: Zero locks — anything can be accidentally deleted

📖 https://learn.microsoft.com/en-us/azure/azure-resource-manager/management/lock-resources


💰 CATEGORY 3 — Cost Controls (15 points)

Detects: "surprise bill" + "VMs running in dev all weekend"

Check 3.1 — Budget alerts configured (8 pts)

az rest --method get \
  --url "https://management.azure.com/subscriptions/<sub-id>/providers/Microsoft.Consumption/budgets?api-version=2023-11-01"

Scoring:

  • 8 pts: Budget exists with notifications at 80% and 100%
  • 4 pts: Budget exists but missing thresholds or notifications
  • 0 pts: No budgets — first warning is the invoice

📖 https://learn.microsoft.com/en-us/azure/cost-management-billing/costs/tutorial-acm-create-budgets

Check 3.2 — Azure Advisor cost recommendations addressed (4 pts)

az advisor recommendation list --category Cost \
  --query "[?impact=='High'].{problem:shortDescription.problem}" \
  -o table

Scoring:

  • 4 pts: Zero high-impact cost recommendations pending
  • 2 pts: 1–3 high-impact recommendations
  • 0 pts: 4+ unaddressed — systematic waste

📖 https://learn.microsoft.com/en-us/azure/advisor/advisor-cost-recommendations

Check 3.3 — No cost-leaking VMs (3 pts)

az vm list -d --subscription <sub-id> \
  --query "[?powerState=='VM stopped'].{name:name, rg:resourceGroup}" \
  -o table

Note: "VM stopped" (not deallocated) still incurs compute charges.

Scoring:

  • 3 pts: No VMs in stopped-but-allocated state
  • 0 pts: VMs burning money while stopped

📖 https://learn.microsoft.com/en-us/azure/virtual-machines/states-billing


🔐 CATEGORY 4 — Identity & Access (20 points)

Detects: "everyone is Owner" + "secrets in code instead of Managed Identity"

Check 4.1 — Owner count ≤ 3 per subscription (6 pts)

az role assignment list --role Owner \
  --subscription <sub-id> \
  --query "[].{principal:principalName, type:principalType}" \
  -o table

Scoring:

  • 6 pts: ≤ 3 Owners at subscription level
  • 3 pts: 4–5 Owners
  • 0 pts: 6+ Owners — the "everyone is Owner for convenience" anti-pattern

📖 https://learn.microsoft.com/en-us/azure/role-based-access-control/best-practices

Check 4.2 — Managed Identities in use (7 pts)

az identity list --subscription <sub-id> \
  --query "[].{name:name, rg:resourceGroup}" \
  -o table
az vm list --subscription <sub-id> \
  --query "[].{name:name, identityType:identity.type}" \
  -o table

Scoring:

  • 7 pts: User-Assigned or System-Assigned MI actively used on resources
  • 3 pts: Some resources have MI, key resources don't
  • 0 pts: No Managed Identities — likely using keys/secrets everywhere

📖 https://learn.microsoft.com/en-us/entra/identity/managed-identities-azure-resources/overview

Check 4.3 — Defender for Cloud enabled (7 pts)

az security pricing list \
  --query "value[?name=='CloudPosture' || name=='VirtualMachines' || name=='StorageAccounts'].{service:name, tier:pricingTier}" \
  -o table

Scoring:

  • 7 pts: CSPM + at least one workload protection (Standard tier)
  • 4 pts: Only free tier CSPM
  • 0 pts: Defender not configured — zero security visibility

📖 https://learn.microsoft.com/en-us/azure/defender-for-cloud/get-started


🌐 CATEGORY 5 — Production Web Protection (15 points)

Detects: "startup running prod on .azurewebsites.net with no WAF"

Check 5.1 — Custom domain on production web apps (5 pts)

az webapp list --subscription <sub-id> \
  --query "[].{name:name, defaultHost:defaultHostName, customDomains:hostNames}" \
  -o json
az staticwebapp list --subscription <sub-id> \
  --query "[].{name:name, defaultHost:defaultHostname, customDomains:customDomains}" \
  -o json

Scoring:

  • 5 pts: Production apps use custom domains
  • 2 pts: Mix of custom and default domains
  • 0 pts: Production on .azurewebsites.net / .azurestaticapps.net
  • N/A: No web apps (score 5 pts — not applicable)

📖 https://learn.microsoft.com/en-us/azure/app-service/app-service-web-tutorial-custom-domain

Check 5.2 — WAF or Front Door protecting web endpoints (5 pts)

az network application-gateway waf-policy list \
  --subscription <sub-id> \
  --query "[].{name:name, rg:resourceGroup}" \
  -o table
az afd profile list --subscription <sub-id> \
  --query "[].{name:name, sku:sku.name}" \
  -o table

Scoring:

  • 5 pts: WAF policy active OR Front Door Premium/Standard protecting endpoints
  • 2 pts: Front Door exists but without WAF enabled
  • 0 pts: Web apps directly exposed without any L7 protection
  • N/A: No public web apps (score 5 pts)

📖 https://learn.microsoft.com/en-us/azure/web-application-firewall/overview

Check 5.3 — DDoS protection stance (5 pts)

az network ddos-protection list \
  --subscription <sub-id> \
  --query "[].{name:name, rg:resourceGroup}" \
  -o table

Scoring:

  • 5 pts: DDoS plan active on VNets with public IPs, OR architecture is PaaS-only (built-in protection)
  • 0 pts: VMs/IPs exposed without DDoS plan

Note: Pure PaaS architectures (App Service, Static Web Apps, Container Apps without VNet) get Azure's built-in DDoS at the platform level. Score 5 if no VNets with public IPs exist.

📖 https://learn.microsoft.com/en-us/azure/ddos-protection/ddos-protection-overview


📋 CATEGORY 6 — Operational Foundations (10 points)

Check 6.1 — Activity Log forwarding to Log Analytics (5 pts)

az monitor diagnostic-settings subscription list \
  --subscription <sub-id> \
  -o json

Scoring:

  • 5 pts: Activity Logs forwarded to Log Analytics workspace
  • 2 pts: Forwarded to storage only (limited query capability)
  • 0 pts: No diagnostic settings — no audit trail

📖 https://learn.microsoft.com/en-us/azure/azure-monitor/essentials/activity-log

Check 6.2 — Backup exists for stateful resources (5 pts)

az backup vault list --subscription <sub-id> \
  --query "[].{name:name, rg:resourceGroup}" \
  -o table

Scoring:

  • 5 pts: Recovery Services vault with active backup items
  • 2 pts: Vault exists but empty (no backup items)
  • 0 pts: No backup — one deletion away from data loss

📖 https://learn.microsoft.com/en-us/azure/backup/backup-overview


Accepted exceptions (optional)

If the user provides a list of accepted exceptions, do not flag those items. Instead, note them in the report as Accepted Exception with the reason provided.

Example format the user may provide:

CheckReason
2.1 Management GroupsSingle subscription by design, MGs planned for Series B
3.1 Budget alertsCost alerts handled by external FinOps tool (Kubecost)
5.3 DDoS protectionPaaS-only architecture with no VNet-attached public IPs

When exceptions are provided:

  • Skip the flagged checks in scoring
  • List them in a separate "Accepted Exceptions" section at the end of the report
  • Recalculate the overall score excluding excepted checks

Expected output

Report header (mandatory — use this exact format)

Digital Native Governance Check Report

FieldValue
ScoreXX / 100
Level🟡 Developing
Subscription(name + ID)
Assessment DateYYYY-MM-DD

Score breakdown

CategoryScoreMaxStatus
🏥 Alerting & ObservabilityX20🟢/🟡/🔴
🏗️ Subscription TopologyX20🟢/🟡/🔴
💰 Cost ControlsX15🟢/🟡/🔴
🔐 Identity & AccessX20🟢/🟡/🔴
🌐 Web ProtectionX15🟢/🟡/🔴
📋 Operational FoundationsX10🟢/🟡/🔴
TOTALXX100Level

Category status: 🟢 ≥ 80% | 🟡 40–79% | 🔴 < 40%

🚨 Critical gaps (blocks enterprise deals)

Items that would immediately fail a security questionnaire. List with severity.

⚡ Quick wins (fix this week)

Top 3 highest point-gain items with lowest effort. Include the az command to fix.

📈 Roadmap to next level

What to tackle next month to reach the next governance maturity level.

Remediation guidance

For each ❌ or ⚠️ finding, include in the output:

  1. The specific az CLI command to remediate (suggest only — do not execute)
  2. Use GetAzCliHelp to validate the command syntax before suggesting
  3. Include the 📖 documentation link from the check that failed

📚 Learn more

Link each finding to the relevant "Digital Natives on Azure" article for deep context.

Sample output

The following is a redacted example of what the report looks like when run against a subscription.

Digital Native Governance Check Report

FieldValue
Score52 / 100
Level🟡 Developing
Subscriptioncontoso-prod-001 (a1b2c3d4-e5f6-7890-abcd-ef1234567890)
Assessment Date2026-07-15

Score breakdown

CategoryScoreMaxStatus
🏥 Alerting & Observability1020🟡
🏗️ Subscription Topology420🔴
💰 Cost Controls1215🟢
🔐 Identity & Access1320🟡
🌐 Web Protection715🟡
📋 Operational Foundations710🟡
TOTAL52100🟡 Developing

🚨 Critical gaps (sample)

#FindingCategoryPoints Lost
1No Management Groups — flat subscription topology🏗️ Topology-8
2Single subscription for all environments🏗️ Topology-8
35 Owners at subscription level🔐 Identity-3
4No Resource Health alerts configured🏥 Alerting-6

⚡ Quick wins (sample)

ActionPointsEffortCommand
Create Service Health alert for all 4 event types+410 minSee below
Add Resource Health alerts+610 minSee below
Configure budget with 80%/100% notifications+45 minSee below

Remediation guidance (sample)

# Create budget with alert notifications
az rest --method put \
  --url "https://management.azure.com/subscriptions/a1b2c3d4-.../providers/Microsoft.Consumption/budgets/monthly-budget?api-version=2023-11-01" \
  --body '{"properties":{"amount":15000,"category":"Cost","timeGrain":"Monthly","timePeriod":{"startDate":"2026-07-01"},"notifications":{"at80":{"enabled":true,"operator":"GreaterThanOrEqualTo","threshold":80,"contactEmails":["ops@contoso.com"]},"at100":{"enabled":true,"operator":"GreaterThanOrEqualTo","threshold":100,"contactEmails":["ops@contoso.com"]}}}}'

# Remove excess Owner assignment
az role assignment delete --assignee user@contoso.com --role Owner --scope /subscriptions/a1b2c3d4-...

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.