Connection Health

Connection state describes configuration and provider access; sync state describes a single execution. Team issues aggregate the failures that need user or operator attention. Show all three at the appropriate level.

Connection Checks

For each connection from GET /connections, monitor:

• status: whether the connection is active;
• app: the selected app, including its type category;
• isErrorState: whether the most recent fetch failed;
• currentSync: the latest sync execution's status, updatedAt, and message;
• accountingStartAt / accountingEndAt: the accounting window for PMS connections;
• enabledFlows and the app's importers / extractors: enabled flows and source capabilities; and
• related structured issue context from GET /teams/issues, which also surfaces credential, reconnect, and outdated-connection conditions.

Do not assume an accepted connect or sync request means provider work is complete. Poll the returned sync or connection resource where documented, or use an event-driven flow when a public event exists.

Partner Portfolio Inventory

Use List partner connections (GET /partner/connections) to build a partner-level connection inventory in one regional partition. Each row is a managed VRI or VRT team and includes active user-visible connections, app identity and category, isErrorState, and nullable currentSync. Teams without visible connections remain in the response.

Search matches team and company names. Lifecycle, app, and isErrorState filters apply before pagination.total and page selection. An app filter selects teams with an active connection to that app. isErrorState=true selects teams with a failed active connection; false selects teams without one. When combined, error state applies to the selected app. These filters do not remove the teams' other connections from the response. Sort by name or createdAt; prefix the value with - for descending order. The default is sort=-createdAt. For a multi-region portfolio, call each region advertised by GET /me and retain region identity when combining results.

Repair Actions

Use issue code and context to choose a specific action:

• reconnect a broken connection — POST /connections/{id}/connect or retry its sync — POST /connections/{id}/sync;
• update settings — PUT /connections/{id} before reconnecting an outdated connection;
• configure a missing PMS accounting start; or
• correct source data that causes repeated item failures.

After repair, re-read connection state and the team issues snapshot. Do not hide an issue only because one retry request returned success.

Related

• Connections
• Portfolio health