Status
Live monitoring is not connected. This page will not pretend it is.
There is no status feed, no uptime probe, and no incident history to render. A green “all systems operational” bar would be a statement about production we cannot make.
Live feedNot connected
Incident historyNone published
Components named7
01 — This page
Do not read the absence of incidents as uptime.
An empty incident list on a page with no provider is not evidence that anything is up. It is evidence that we have not wired the feed.
02 — Components / 7 components
What a later feed will have to name.
These are the surfaces a public status page is for. Each row says what it is and that we are not reporting on it today.
- Marketing site
- This Next.js app — catalog, pricing, docs, and the pages you are on. No probe reports its uptime here.
- App and dashboard
- Signed-in workspace, product settings, billing views. Same gap: nothing on this page is reading those routes.
- Auth
- Sign-in, signup, and invite flows. A later status feed would list this as its own component, not as a green banner.
- Checkout
- Checkout in this build records a waitlist and takes no payment. There is no payment processor to report.
- Product runs
- Hosted and connected tools, agents, and workflows. Run success and queue depth are not exported to a public status API yet.
- Model providers
- Inference is a subprocessor. Their outages are theirs. Credits for availability apply only when we are down — see Security.
- Customer systems
- QuickBooks, HubSpot, mailboxes, and the rest. Connected products act through OAuth you can revoke. Their status is not ours to paint green.
03 — Source
Where the numbers will come from, when they exist.
Spec 09 wants this route to proxy a status provider. Until that contract is live, the honest output is this explanation.
The feed
- Intended source
- A status-provider API, proxied by this route (spec 09). No provider is configured, so there is nothing to proxy.
- What would land here
- Component health, incident open/close times, and a history a buyer can check without asking sales.
- What would not land here
- Security breaches. Those go to the workspace owner, not first to this page and not first to social media.
- What you can check today
- Availability SLAs, RPO, and RTO by plan on /security. Starter and Growth have no availability SLA.
Until then
If the marketing site itself fails to load, you will not see this page either. That is the whole current signal. We will not invent a 30-day uptime percentage, a status-dot colour, or a historical incident we did not record.
When a provider is connected, this route stays SSR, the inventory above becomes the component list, and incidents get timestamps rather than adjectives.
04 — Commitments / 4 plans
Availability promises live on Security, not on a fake meter.
RPO, RTO, and SLA by plan, copied from the same table Security and Trust render. Credits apply on Scale and Enterprise when we miss the published number — not on Starter or Growth.
| Plan | RPO | RTO | Availability SLA |
|---|---|---|---|
| Starter | 24 hours | 8 hours, business hours | None. Best effort, no service credits |
| Growth | 24 hours | 8 hours, business hours | None. Support response SLA only |
| Scale | 1 hour | 4 hours | 99.9% monthly |
| Enterprise | 15 minutes | 1 hour | 99.95% monthly, with credits |
Self-hosted products run in your environment. Their uptime is yours. Model-provider outages are passed through; we credit only when we are down.
Questions
Need the contractual version, not a dashboard?
Availability language, credits, and exclusions are on Security and in the SLA. Demo is the path if your MSA needs a human.