[Epic] Diagnose and resolve Azure failures using live evidence

## Problem statement

Troubleshooting responses often stop at generic checklists or isolated log queries. Users need evidence collection, diagnosis, remediation, and verification connected into one workflow.

## Outcome

Users can diagnose Azure application and infrastructure failures from live operational evidence and verify the remediation.

## Goals

- Collect relevant logs, metrics, deployment history, health, access, and configuration evidence
- Identify likely root causes and distinguish uncertainty
- Recommend safe, ordered remediation
- Verify service and application health afterward

## Non-goals

- Claiming root cause without evidence
- Applying destructive remediation without confirmation

## Success criteria

- [ ] Priority failure modes have evidence-driven workflows
- [ ] Diagnostics produce actionable findings rather than portal checklists
- [ ] Remediation completion includes post-change verification

## Dependencies

Azure Monitor, AppLens, resource health, service diagnostics, deployment history, and RBAC.

Start by reviewing the listed dependencies: Azure Monitor, AppLens, resource health, service diagnostics, deployment history, and RBAC. Define priority failure-mode workflows that collect live evidence, state uncertainty, recommend safe remediation, and verify health afterward. Done means the success criteria are met without claiming unsupported root causes or applying destructive changes without confirmation.

Written by the indexing model from the issue text.