---
title: Status and incidents
description: How SV3 reports service degradation, pauses, and recovery.
---

> **For AI agents:** the complete documentation index is at [llms.txt](/llms.txt). Append `.md` to any page URL for its markdown version.

SV3 separates contract safety from the services used for discovery and
convenience. An API, indexer, chart, image, or keeper outage does not by itself
change on-chain reserve, supply, debt, collateral, or the floor. It can still
make information stale or normal actions difficult to use.

## What status updates identify

During a user-impacting incident, updates should state:

- the UTC time and affected component or market;
- whether the chain and direct contract reads are available;
- the last independently verified block and deployed revision when relevant;
- whether indexed data is stale or unavailable;
- the current global and market pause state;
- which actions are affected; and
- when the next update is expected.

An HTTP health response is not proof that every dependency is ready. A service
may be live while its index checkpoint is stale. The application should label
that state and refresh execution-critical information from the chain.

## Full pause

The protocol is upgradeable and its administrator can indefinitely pause all
state-changing market actions. A full pause can block buys, sells, redemption,
repayment, surrender, and fee claims. It has no automatic expiry.

A pause is a protective control, not proof that an issue is resolved. Resuming
actions requires accountable human authorization after contract, reserve,
deployment, and accounting state are reconciled.

## Honest incident language

Status updates distinguish verified facts from hypotheses. They do not describe
USDC as insured or guaranteed, promise an outcome before investigation, expose
private user or credential data, or call the protocol immutable.