Reviewing & Approving

Pause, Resume & Errors

Productions are long-running and asynchronous — sometimes you need to halt one, sometimes one halts itself. This guide covers both.

Where the controls live

Every state-change control — Pause, Resume, Retry Phase, Approve, Reject — lives in the action bar at the top of the production detail page, rendered dynamically from production.available_actions. Which buttons you see depends on the production's current state and working_state. The phase tabs themselves are mostly viewers.

Pause and Resume

Why you'd pause

  • You spotted something off in Phase 1 and need time to think before approving.
  • A reviewer is unavailable; you don't want the next phase to auto-fire.
  • You started a multi-language batch and want to suspend everything during a service incident.
  • You're about to hit a storage limit and need to free space first.

How to pause

When the production is in progress, the action bar at the top has a Pause button.

Clicking it:

  • Stops the next phase from firing.
  • Holds the production in its current state with a yellow suspended banner: "Production paused at Phase X. Next phase will not start automatically. Use Resume to continue."
  • Does not roll back any work already done.

How to resume

Once paused, the action bar shows a Resume button. Click and the production picks up exactly where it left off.

Heads up — quota is not refunded by pausing. A paused production still occupies its monthly video slot.

The Errors tab

The Errors tab on the production detail page only appears when the production has at least one error in its history. If you've never seen this tab on a production, that production has been clean.

What you'll find inside

  • Per-phase error messages for any phase that failed (Phase 1 through Phase 5).
  • The full error JSON if available — error type, message, stack trace, n8n execution ID, AI service response.
  • Timestamps for when each error occurred.

The errors are displayed read-only. There's no manual Retry button on the user UI — retries are scheduled automatically (see below).

How error recovery actually works

Studio Cut Video does not silently fail. When a phase errors:

  1. Capture. The full error context is stored against the failing phase.
  2. Visual signal. A red banner appears on the production page: "Phase X failed. Next phase will not start automatically. Use Retry Phase to try again." The Errors tab appears.
  3. Auto-retry. A cron job may re-attempt the failed phase. Behaviour varies by error class.
  4. Manual retry. The action bar surfaces a Retry Phase button — you can trigger an immediate retry yourself without waiting for the auto-retry cycle.
  5. Backoff. If retries keep failing, the production stops attempting silently and waits for your input.

What you, the end user, can do

  • Click Retry Phase. It's right there in the action bar when a phase has errored. Most transient failures (rate limits, network blips, AI API hiccups) resolve on a retry.
  • Pause and inspect. If you want to investigate before retrying, pause the production first.
  • Reject and regenerate from an earlier phase. If the error is content-related (e.g. AI keeps refusing your script for safety reasons), reject from a working earlier phase and reword.
  • Edit Production and retry. If the error traces back to a setting (a font that the renderer can't find, a voice that doesn't support your language), open the Edit Production modal, fix the setting, then click Retry Phase.
  • Contact support. If retries have exhausted and the error is environmental (storage, plan limits, integration auth), reach out.

Real scenarios

Scenario A — Transient AI rate limit

Phase 2 fails with "Rate limit exceeded — Gemini API". The red banner shows the error and the Retry Phase button appears in the action bar.

  1. Wait a minute or two for the rate limit to clear.
  2. Click Retry Phase in the action bar.
  3. Phase 2 re-runs and succeeds.

Scenario B — Storage exhausted before Phase 1

Phase 1 refuses to start. A red banner: "Insufficient storage — at least 200 MB free required."

  1. Open the Asset Library.
  2. Bulk-select unused intro footage and move to Trash.
  3. Open Trash → Empty Trash (items in trash still count toward usage).
  4. Return to the production and click Resume.

Scenario C — Phase 4 render fails repeatedly

Render service has a persistent issue with one specific font in your brand kit. Auto-retry exhausted; production is stuck.

  1. Pause the production.
  2. Open Edit Production, change the Body Font to a different option, save.
  3. Resume — Phase 4 re-runs with the new font and succeeds.
  4. Update your Brand Library preset to drop the problem font.

Scenario D — Wrong content half-way through

You realise Phase 2 generated assets for the wrong audience.

  1. Pause.
  2. Open Edit Production, change Target audience, save.
  3. Reject Phase 2 (in the Assets tab) with feedback: "Re-target this audience."
  4. Resume — assets regenerate.

Tips & gotchas

  • Pause is not Cancel. There's no destructive cancel for end users — pausing keeps your work and quota slot. If you genuinely don't want the production, leave it paused; it won't auto-advance.
  • Auto-Approve does not bypass errors. Even with Auto-Approve on, an errored phase stops the production.
  • The Errors tab persists per-phase errors. It shows whatever is in production.phaseN_error for each phase plus the workflow error_json. It is read-only and historical.
  • Retry Phase is your friend. It's the explicit, visible manual retry button. Use it before contacting support.
  • Manager controls exist but are admin-only. If you're an agency owner or white-label admin, you may see additional phase-jump controls; standard end users won't.

See also