How-to guides
These recipes explain how to complete real tasks in Azure Support Agent. Each procedure identifies the application route, permissions and prerequisites, numbered actions, expected result, verification, safety or rollback considerations, and troubleshooting.
| Area | Recipes |
|---|---|
| Core and workload operations | Dashboard, Chat, Deep Investigation, Proactive Support, Monitor, Stats, Workloads, Autopilot, groups, overlaps, and Mission Control |
| Design and assessment operations | Insight Packs, Architectures, Know-Me, Ownership, Estate Graph, Assessments, Performance Profiler, and FMEA |
| Coverage operations | Monitoring, Alerts Manager, Telemetry, Backup/DR, and Connection Capability |
| Estate intelligence operations | Inventory, Tag Intelligence, and Change Explorer |
| Governance and identity | Azure Policy, Identity, PIM, app registrations, and RBAC |
| Lifecycle and investigation | Retirement Radar, reservations, quota, telemetry, evidence, and cases |
| Automations and connectors | Scheduled Tasks, Workbooks, Playbooks, Notifications, and every implemented connector |
| Administration tasks | Providers, tenants, access, security, references, usage, audit, MCP tools, backup, and demo data |
How to use a recipe
- Confirm the route and scope before running a scan or editing a record.
- Check both product permissions and Azure/Graph permissions.
- Review freshness, cache, truncation, and partial-result indicators.
- Verify generated or AI-authored content against source evidence.
- Preview write operations, preserve approvals, and understand rollback before apply.
- Re-query the owning system after a change and preserve verification evidence.
Examples intentionally contain no live tenant identifiers, resource IDs, tokens, receiver addresses, or credentials. Keep operational exports and screenshots out of public documentation unless they are sanitized.