Static Web Apps: all deployments fail "An unknown exception has occurred" at upload phase — East US 2 slice c2gp-slice-0-eus2-0-ase

## Summary
All deployments to our Azure Static Web Apps in **East US 2** fail during the **upload phase** with the generic `An unknown exception has occurred`, and **no server-side build record is created**. Reproduced across three independent environments and against a brand-new clean resource. Production sites remain healthy (serving last-good content) — only *new* deployments fail.

## Deployment slice (common to all failures)
- Slice: **`c2gp-slice-0-eus2-0-ase`** (IP `20.96.255.85`, port 443)
- TLS cert: `CN=*.c2gp-slice-0-eus2-0-ase.p.azurewebsites.net`, issuer *Microsoft Azure RSA TLS Issuing CA 03* (valid)
- Region: **East US 2** · SKU: Standard · _(subscription/resource IDs available privately on request)_

## Reproduced identically in 3 environments, 2 resources
- Local machine (Linux/WSL) **and** Windows client
- **Azure Container Instance** in East US 2 (separate Azure-backbone network)
- **GitHub Actions** runner (official `Azure/static-web-apps-deploy@v1`)
- Against an existing resource **and** a brand-new clean Standard SWA
- Minimal one-file `index.html` payload
- SWA CLI **v1.1.10 and v2.0.9** (StaticSitesClient `689a6c1f…`) — both fail

## What works (rules out client / network / token)
- Token validation **succeeds** (a bogus token yields a distinct `deployment_token provided was invalid`)
- TCP **ESTABLISHED** to the slice; TLS handshake completes (valid MS cert); `curl` returns HTTP 404 at slice root
- No proxy (Windows WinHTTP = Direct, no proxy env vars)
- Generic Azure Blob Storage read/write works fine from the same machine

## Symptom detail
Fails right after `Preparing deployment… DeploymentId: `. The Windows client (verbose) reaches `Determined default file to be: index.html`, then throws. The inner exception is swallowed by the client's top-level handler. The portal's "Retrieve deployment error messages" shows nothing (no build record). Affected env stuck in **Uploading**; the new clean resource stuck in **WaitingForDeployment**.

### Sample failed Deployment IDs
- `936618cb-e872-45e6-8e35-f5de9b91dccb` (CLI v1.1.10)
- `cdf7b2fa-426d-4373-88f4-ca43cd354a6d` (CLI v2.0.9)
- `2cc377ef-8425-4f5e-8173-aeb6d7df9a13` (GitHub Actions)

## Ask
1. Investigate the East US 2 deployment slice/backend `c2gp-slice-0-eus2-0-ase`.
2. Surface the swallowed inner exception behind the generic CLI message.
3. Confirm when deployments through this slice are healthy again.

Happy to provide subscription/resource IDs privately to a Microsoft engineer.