AuditClaw Azure

Azure compliance evidence collection for auditclaw-grc. 12 read-only checks across storage, NSG, Key Vault, SQL, compute, App Service, and Defender for Cloud.

Companion skill for auditclaw-grc. Collects compliance evidence from Azure subscriptions using read-only API calls.

12 checks | Reader + Security Reader roles only | Evidence stored in shared GRC database

Security Model

• Read-only access: Requires only Reader + Security Reader roles (subscription-level). No write/modify permissions.
• Credentials: Uses DefaultAzureCredential (service principal env vars, az login, or managed identity). No credentials stored by this skill.
• Dependencies: Azure SDK packages (all pinned in requirements.txt)
• Data flow: Check results stored as evidence in ~/.openclaw/grc/compliance.sqlite via auditclaw-grc

Prerequisites

• Azure credentials configured (service principal or az login)
• pip install -r scripts/requirements.txt
• auditclaw-grc skill installed and initialized

Commands

• "Run Azure evidence sweep": Run all checks, store results in GRC database
• "Check Azure storage security": Run storage-specific checks
• "Check Azure network security": Run NSG checks
• "Check Azure Key Vault": Run Key Vault checks
• "Check Azure SQL compliance": Run SQL Server checks
• "Check Azure VM encryption": Run compute checks
• "Check Azure App Service": Run App Service checks
• "Check Azure Defender": Run Defender for Cloud checks
• "Show Azure integration health": Last sync, errors, evidence count

Usage

All evidence is stored in the shared GRC database at ~/.openclaw/grc/compliance.sqlite via the auditclaw-grc skill's db_query.py script.

To run a full evidence sweep:

python3 scripts/azure_evidence.py --db-path ~/.openclaw/grc/compliance.sqlite --all

To run specific checks:

python3 scripts/azure_evidence.py --db-path ~/.openclaw/grc/compliance.sqlite --checks storage,network,keyvault

To list available checks:

python3 scripts/azure_evidence.py --list-checks

Check Categories (7 files, 12 findings)

Check	What It Verifies
storage	HTTPS-only transfer, TLS 1.2+, public blob access disabled, network default deny
network	NSG no unrestricted SSH (port 22), no unrestricted RDP (port 3389)
keyvault	Soft delete + purge protection enabled
sql	Server auditing enabled, TDE encryption on all databases
compute	VM disk encryption (encryption at host)
appservice	HTTPS-only + TLS 1.2+
defender	Defender plans enabled (Standard tier) for critical resource types

Authentication

Uses DefaultAzureCredential from azure-identity. Supports:

• Service principal: AZURE_CLIENT_ID + AZURE_TENANT_ID + AZURE_CLIENT_SECRET
• Azure CLI: az login
• Managed identity (when running in Azure)

Minimum roles: Reader + Security Reader (subscription-level)

Evidence Storage

Each check produces evidence items stored with:

• source: "azure"
• type: "automated"
• control_id: Mapped to relevant SOC2/ISO/HIPAA controls
• description: Human-readable finding summary
• file_content: JSON details of the check result

Setup Guide

When a user asks to set up Azure integration, guide them through these steps:

Step 1: Create Service Principal

az ad sp create-for-rbac --name auditclaw-scanner --role Reader --scopes /subscriptions/<SUBSCRIPTION_ID>

Step 2: Add Security Reader Role

az role assignment create --assignee <APP_ID> --role "Security Reader" --scope /subscriptions/<SUBSCRIPTION_ID>

Only 2 roles needed: Reader + Security Reader (subscription-level).

Step 3: Configure Credentials

Set environment variables from the service principal output:

• AZURE_CLIENT_ID (appId)
• AZURE_CLIENT_SECRET (password)
• AZURE_TENANT_ID (tenant)
• AZURE_SUBSCRIPTION_ID

Step 4: Verify Connection

Run: python3 {baseDir}/scripts/azure_evidence.py --test-connection

The exact roles are documented in scripts/azure-roles.json. Show with: python3 {baseDir}/../auditclaw-grc/scripts/db_query.py --action show-policy --provider azure