Research failed
Retry research, fix inputs, and resolve usage limits when a run fails.
Overview
Research runs can fail when inputs are incomplete, sources are unreachable, usage pools are exhausted, or a transient provider error occurs. Most issues are fixable without support by narrowing the question, checking billing, and retrying.
When to use this
- A task shows failed or empty research output.
- Live progress stalled without completing.
- You saw an error about credits, tokens, or plan access.
Steps
Open the task or research panel and read the error message (timeout, usage, validation).
Check inputs: address spelled correctly, municipality selected, custom prompt specific enough, focus links reachable.
Check usage: Settings → Billing → Subscription for remaining research runs and AI tokens; owners see Credits & AI usage. Free workspaces may need pay-as-you-go or upgrade—see Credits and AI usage.
Retry the run. Transient network errors often succeed on a second attempt.
If output is partial, verify fields with sources and citations and add overrides for known fixes.
If failures repeat, email support@toolblocks.app with task link, workflow type, and time (UTC).
Quick diagnosis
| Error pattern | What to try |
|---|---|
| Usage / quota | Upgrade, wait for monthly reset, or reduce batch runs |
| Missing address or jurisdiction | Complete required fields; use autocomplete |
| No sources returned | Narrow prompt; add focus links; treat as uncertain |
| Plan gate | Confirm Team + Revenue seat for full research pools |
Research outputs are draft for review. A “successful” run can still include uncertain fields—always verify citations before client use.