swa deploy fails with "missing pull request id" / "problems communicating with content server" — no CI/CD, plain CLI deploy to Production

Question

I have an Azure Static Web App (no linked GitHub/DevOps repo — deployed purely via the SWA CLI with a deployment token, provider shows as "SwaCli") that has stopped accepting deployments entirely. Looking for anyone who's seen this specific failure mode, or ideas on what else to check before I wait on a support ticket.

Setup:

• Static Web App, Production environment only (no preview/PR environments configured)
• Deploying via: swa deploy ./dist --app-name <app> --env production --deployment-token <token>
• Tested on SWA CLI 2.0.9 and 2.0.10 — identical behavior on both

What happens:

Every deploy attempt fails with the CLI's generic "The deployment binary exited with code 1" message. Running the same command with --dry-run surfaces what's actually happening server-side:

"The content server has rejected the request with: BadRequest

Reason: Request is missing the pull request id."

This happens even though --env production is always explicitly passed, and there's no PR or branch involved anywhere in this deployment — it's a bare CLI push to Production.

After resetting the deployment token via the Portal ("Manage deployment token" → "Reset token"), that specific error changed to a different, more generic one, which has now recurred consistently on every attempt since:

"We are currently experiencing problems communicating with our content server.

Please try again later or file an issue if this behavior continues."

Also worth noting: az staticwebapp environment list shows the Production environment repeatedly getting stuck in "Uploading" status after failed attempts, sometimes eventually settling to "Failed" — never "Ready." The Portal's own Environments page for this resource has also intermittently failed to load entirely.

**
What I've already ruled out:**

• Not a CLI bug — identical behavior across two CLI versions, and a fully re-downloaded deployment binary (cleared ~/.swa/deploy cache)
• Not a spend/budget cap — no Azure Cost Management budgets configured at all, subscription state is "Enabled"
• Not a broader outage — status.azure.com shows no active incidents for Static Web Apps/CDN/Front Door; this appears scoped to just this one resource

Has anyone run into this exact "missing pull request id" error on a non-GitHub, CLI-only production deploy? Or know of anything else worth checking (resource recreation, specific region issues, etc.) before this gets escalated further?