The connection detail page
Clicking a connected card opens that one source's own page: when Scout last read it, what those reads produced, whether the points it raised were worth having, and every control for pausing, reconnecting or removing it.
Screenshot: a connection's detail page.
The header
The line under the name says how the connection is owned and who set it up — for example "ORG CONNECTION · CONNECTED 14 MAY 2026 BY MEGHAN", or, for a mailbox connection, "PER-USER CONNECTION · THE ORG OWNS IT · EACH PERSON AUTHORIZES IT".
Beside it is the authorization badge. It is the one thing to read first:
| Badge | Meaning |
|---|---|
| AUTHORIZED | Everything is granted and working. |
| AWAITING FIRST RUN | Connected, but Scout has not completed a read yet. |
| ONE GRANT EXPIRED | A per-person connection where one person's authorization lapsed. Everyone else is fine. |
| DEGRADED · STALE | The authorization is valid but the data is older than the cadence allows. |
| NEEDS ATTENTION | Something is wrong that has not stopped the connection dead. |
| RE-AUTHORIZE | The provider rejected Scout's authorization. Reconnect it. |
| SYNC FAILED | Reads are failing. The reason and date are on the card. |
Two buttons sit next to the badge: Sync now, which asks for a read immediately instead of waiting for the schedule, and Re-authorize, which starts the connect flow again for this exact source.
Run history
A table of recent reads with five columns: STARTED, TOOK, SEEN (records the source offered), NEW (records Scout had not seen before) and RESULT. A result is either OK or the reason it was not — and on a per-person connection it names the person, for example "PRIYA · 401".
This is the page to check when someone asks whether a connection is really working. A row with a result is proof that Scout asked. No rows at all means no read happened — not that a read happened and found nothing.
What Scout did with it · last 30 days
Three tiles: POINTS RAISED, ACTED ON and MARKED WRONG. The page says why they are there: a feed that raises points nobody acts on is costing attention, and this is the argument for keeping a connection — or dropping it.
Freshness
How long it has been since the last successful read, the cadence, and the point at which Scout treats the data as stale — "Every 15 min · Stale after 60 min" for a mailbox, "Nightly 02:00 · Stale after 10:00 local" for accounting. Freshness is wall-clock rather than elapsed time, which is the honest measure for a nightly feed: what matters is whether the numbers arrived before you needed them.
The settings on the page
- Look-back — how far into history Scout reads: 1 month, 6 months, 1 year, 2 years, 3 years, 4 years or 5 years. Five years is the cap. Mailboxes default to 1 year and accounting to 2 years.
- Degraded alerts — when Scout tells you this connection is in trouble. Three choices: Immediate ("Email + in-app the moment it trips"), With the morning brief (the default for catalogue feeds), and In-app only (shown in Scout, nothing by email).
- When someone leaves — on a per-person connection, what happens to a departing colleague's mail evidence: Keep all or Purge content. It is set once per workspace and shown to each person at the moment they authorize, so nobody learns it at offboarding.
- Cadence · field map · pause — an expander holding the rest: cadence and field mapping, Pause without disconnecting, and Disconnect.
Screenshot: the lifecycle controls.
The five lifecycle actions, and what each does to your data
Each one asks you to confirm, and the confirmation says exactly what will happen.
| Action | Label you will see | What it does |
|---|---|---|
| Reconnect | Reconnect provider authorization, or Retry provider authorization if the first attempt never finished | Opens the provider's authorization flow again for this exact source and replaces the stored credential. It can never bind a different mailbox or company to an existing connection. For a named shared mailbox it repeats the configuration instead of a browser sign-in. |
| Pause | Pause reads (or Pause actions for a connection that sends you notifications rather than reading records) | Stops new reads and keeps everything else. Work in progress stays resumable, and the connection keeps its claim on the addresses it covers, so a conversation does not suddenly produce duplicate items from another mailbox. |
| Resume | Resume reads / Resume actions | Re-checks the live authorization, then starts reading again from where it stopped. |
| Revoke | Disconnect and release | Fences off access, releases the addresses this source covered, and queues deletion of the stored credential — revoked at the provider where that is possible, destroyed locally where it is not. Existing Scout items remain. |
| Retire | Remove permanently | Only offered after a disconnect, and terminal. Scout finishes the cleanup and removes the source from Connections. Items it already raised are kept. |
Nothing here deletes your work. Points, receipts and the value ledger survive a disconnect, because they are Scout's record of what happened, not a copy of the provider's records.
Good to know
- A personal mailbox can only be changed by the person who owns it: "Only the mailbox owner can change this personal source." An Owner or Admin can still inspect its authorization trail.
- A retired source cannot come back to life with a late answer. Every read re-checks that the source is still active.
- Disconnecting a Microsoft source destroys Scout's copy of the credential, but the Microsoft consent and any Exchange rules stay until a tenant administrator removes them. Scout says so rather than implying it cleaned up for you.
- Sync now does not create data out of nothing. If there is nothing new upstream, the run history records an honest empty read.