Common issues and how to fix them

The four most common support tickets, with the self-serve fix for each. If the fix below does not resolve the issue, hit the floating "?" in the bottom-right of any page to open a ticket.

My Brief didn't arrive

Briefs are generated by a job that runs once a day at 11:00 UTC. There is no per-workspace generation time and no timezone setting for it. If nothing has appeared well after that, walk this checklist:

1. Check /today first. On the Free plan the Brief is in-app only — it is generated for you but never emailed, so an empty inbox is expected and /today is where it lives. Pro and above also receive it by email.
2. Check digest frequency. /settings/notifications → notification cadence. That setting governs which notification emails you receive; the Brief still generates regardless, so check /today directly.
3. Check your tier. Free tier renders the top-3 items only — if the Brief looks short, that is by design. Upgrade to see the full set.
4. Hit "Regenerate." The variant dropdown on /today triggers an on-demand regeneration. Free-tier users — note this counts against the 50-query monthly cap.

If none of the above lands the Brief, open a ticket with the date and your workspace ID.

Search returns nothing

Faceted search runs over the indexed federal-civilian research library. Empty results almost always come from one of three causes:

1. Filters too narrow. Open the filter panel and clear date-range, agency, and source-type filters one at a time. If a filter set returns results when one filter is dropped, that filter was over-restrictive.
2. Research library seed status. /settings/data → Corpus status shows the per-source seed health. If SAM or EDGAR is flagged "stale," that source is being re-ingested and will return to normal within an hour.
3. Query too specific. A 12-word boolean query with multiple exact-match phrases will return zero for most workloads. Drop quotes around the most-specific phrase and re-run.

Saved searches that suddenly return zero where they previously returned hits are usually caused by case #2 — a stale source library. The saved-search detail page surfaces the source-status badge.

Webhook delivery_failed

/settings/webhooks/{endpoint}/deliveries lists every failure with the response code and body. The two modal causes:

1. Signing secret rotated, deployment did not. Rotation has a 24-hour grace window, so deployments have time to swap. If the failure window started 24 hours after rotation, that's the cause — update the secret in your secrets manager and replay the failed deliveries via the "Retry" button.
2. Endpoint returns 4xx for valid payloads. The endpoint owner needs to look at their logs; 4xx is a "this request was malformed" signal that we honor without retry. Check that your endpoint accepts the application/json content-type and the request body shape that the API docs at /docs/api describe.

After 10 consecutive failed deliveries (a running counter — no time window) the endpoint is auto-disabled. Re-enable it from /settings/webhooks.

API 429 rate limit

429 responses include a Retry-After header in seconds. The two modal causes:

1. Single key over per-minute cap. Spread requests across multiple keys, one per integration. The per-key cap is 60/min on Enterprise, 120/min on Government.
2. Workspace over daily cap. 10,000/day on Enterprise, 50,000/day on Government. If you regularly hit the workspace cap, contact sales — Enterprise contracts can lift the daily cap to a custom number.

Polling integrations should respect Retry-After and back off; do not retry tighter than the header suggests. The rate-limit dashboard at /settings/api-keys shows per-key usage and the 5-minute / 1-hour / 24-hour windows.

Still stuck

Open a ticket via the floating "?" button. Tickets carry a five-minute SLA on the data-concern category; other categories are next-business-day.