Reliability and incident policy.
What Signclad measures, what it will publish, and how you will hear about an incident. A dedicated status host is not provisioned yet, so this page is the policy rather than the dashboard.
There is no live dashboard yet.
A dedicated status host is not provisioned. Until it is, report anything that looks broken
through the contact form and we will tell you what we know. The
changelog records what has shipped.
What it will cover
- API
- Availability and latency for reads and creates, measured against the published budgets: p95 under 300 ms for reads and under 800 ms for creates, excluding file conversion.
- Signing
- Whether recipients can open and complete documents.
- Web app
- Whether senders can sign in, prepare, and send.
- Conversion
- DOCX and Office conversion, measured against p95 under 15 seconds for a twenty-page document.
- Webhooks
- Dispatch latency and queue depth, against a budget of under 5 seconds from event to dispatch.
- Delivery through our providers, since a signing link nobody receives is an outage.
Planned incident updates
- Once the status host is available, incidents will be posted there as soon as they are confirmed, with a severity and the components affected, and updated at least hourly until resolved.
- Subscribers will receive email updates. There is no need to watch the page.
- A post-incident review is published for anything that affected sending, signing, or the evidence pipeline.
- Uptime figures are published only for periods our monitoring can substantiate. We will not print a number we cannot show you the measurement for.
- Webhook delivery is at least once and retries for about three and a half days, so a receiver outage on your side is not data loss. Everything missed can be replayed.