Signals & reporting
The core decides locally and reports afterwards. This page is the honest list of what that report contains.
Last updated 17 August 2026
What a signal is
A signal is a short record of a decision the core already made: what kind of event it was, how it was classified, what action was taken, and when. It is a summary of behaviour, not a copy of traffic.
What we receive
- Product identifier — which of your systems reported.
- Event class and severity from the classifier.
- Action taken locally (allowed, throttled, challenged, blocked).
- Timestamp and the core's release version.
- Coarse counters used for your usage rollups.
What we never receive
- Request bodies, form contents or uploaded files.
- Your customers' names, emails or account records.
- Session cookies, tokens or credentials.
- Your database contents. The core's state stays in your database.
This is a design constraint, not a setting. There is no mode that turns it off, because a payload we never hold cannot leak from us.
Where to read them
- My security — live posture, signal volume and open incidents per system.
- Account overview — the short version, plus your onboarding state.
- Support — attach an incident to a ticket when you want us to look at it.
Retention
Individual signals are kept for the operating window needed to investigate incidents, then rolled up into daily aggregates and discarded. Aggregates carry no per-event detail. Closing your contract removes your signal history from our side.
Queueing and back-pressure
If we are unreachable or rate-limiting, the core queues reports locally and retries with backoff. Protection is unaffected while that happens — reporting is downstream of the decision, never upstream of it.
Questions about this page? Ask the assistant inside your account, or open a ticket there — it reaches us directly.