Request log, safe retries and limits
Updated 15 Nov 2026
When an integration misbehaves, you want to see what List79 received and why it answered the way it did. This article covers the request log, the token checker, read-only keys, safe retries and the rate limit.
The request log
Open the product's Keys and portal page and choose Open the request log. Owners and admins can see it. It lists calls from the last 7 days:
- API: every call.
- Widget and client API: changes and failures (not every page load).
Each row shows the time, the call, its status, how long it took and which key made it. Failed calls say why in plain words, for example that a user token has expired or was signed with the wrong secret. Filter by Failed or Token problems to find what's wrong quickly. No request bodies or user details are kept. Test mode has its own log.
Request IDs
Every answer from the API carries a List79-Request-Id header, such as req_3f9a1c27e0b4d561. The same ID is shown under the call's time in the request log. Log it on your side, and quote it if you contact us about a call.
Check a user token
The widget, portal and client API sign people in with a user token your server makes. If people can't vote or comment, paste one of your tokens into Check a user token at the top of the request log. It isn't stored. List79 shows who the token signs in, when it expires and what's in it, or exactly what's wrong, for example:
- it has expired: make a fresh token when the page loads;
- it was signed with a different secret, or with the other mode's secret (live and test each have their own);
- it isn't signed with HS256;
- it has no
subclaim with your user's ID.
Read-only keys
When you create a secret key, you can tick Read-only. A read-only key can read requests, votes and everything else, but any change it tries gets 403 Read-only key. Use one for dashboards, reports, and AI tools that should look without changing anything. Read-only keys are marked on the Keys page.
Retry safely with Idempotency-Key
If a call times out, you can't tell whether it went through. To retry without creating a second request or vote, send an Idempotency-Key header with any change (POST, PATCH, DELETE): a unique value per change, such as a UUID, up to 255 characters.
curl -X POST https://list79.com/api/v1/requests \
-H "Authorization: Bearer sk_…" \
-H "Idempotency-Key: 6f1c2b9e-0d4a-4c1e-9b7a-2f5d8e3a1c40" \
-H "Content-Type: application/json" \
-d '{ "title": "Export to CSV", "user": { "id": "u_123" } }'
- Retry with the same key and the same call within 24 hours, and you get the first answer back without anything happening again. The answer carries
Idempotent-Replayed: true. - The same key with a different call gets
422: use a new key for each change. - If the first call is still running, a retry gets
409; try again in a moment.
The rate limit
Each product can make up to 300 API calls a minute; the MCP server shares this limit. Every answer tells you where you stand:
| Header | Meaning |
|---|---|
RateLimit-Limit | Calls allowed a minute (300). |
RateLimit-Remaining | Calls left in this minute. |
RateLimit-Reset | Seconds until the count starts again. |
Over the limit, calls get 429 Too many requests with a Retry-After header saying how many seconds to wait. Wait that long and retry, with the same Idempotency-Key for changes.