The Health dashboard is the one-stop shop for managing your Bid Library's content. Rather than discovering problems one document at a time, it brings the four things that determine whether your library can be trusted into a single view.

Taken together, these answer one question: can we rely on what is in here today? A library that looks full can still be stale, full of gaps, and inaccessible to the people who need it.

Knowledge gaps are raised automatically while drafts are generated. When BidScript finds that a draft is missing information, evidence or alignment that the tender expects - and your library has nothing to supply it - that gap is recorded here.

This is the most valuable list on the dashboard, because it is evidence-based. It tells you not what someone thinks might be missing, but what BidScript actually could not find while trying to answer a real tender question.

Recurring gaps in the same subject area usually mean one missing document rather than many. Adding it closes every future instance at once.]

See Content gaps and flagged issues for how gaps are raised and resolved inside the Answer Workspace.

The freshness section summarises how current your library is: which documents are Fresh, Ageing or Stale, and which have reviews due or overdue.

Freshness is calculated from each document's Modified date against the threshold set for its library. Reviews you schedule manually override that automatic tag.

Staleness does not remove a document from AI search — a stale document will still be used in drafts if it is the best match. That is exactly why this list matters.

For thresholds, review dates and the full tag definitions, see Bid Library Freshness.

Access requests are raised when someone needs access to library content they cannot currently reach.

They appear here so administrators can review them in one place, rather than fielding requests over email and chat, and grant or decline access accordingly.

A pattern of repeated requests for the same library is usually a permissions problem rather than a series of individual exceptions — worth fixing at the library level instead of request by request.

Flagged items are content issues raised by people rather than detected automatically. They highlight missing, incomplete or out-of-date content, and act as your library's improvement backlog.

The BidBot route matters because it removes the friction. Someone querying BidBot who notices an answer is out of date can raise the issue in the moment, without leaving what they were doing or knowing where the Health dashboard is.

The "why it matters" and "example" fields are what make a flagged item actionable by someone who was not there when it was raised. A flag that only says "we need more social value content" gets ignored; one that names the tender question it failed and the kind of document that would answer it gets closed.

Flagged items can be assigned to a team member, marked resolved once the content has been added, or deleted where they are no longer relevant. Deletion is permanent.

See Flagged Items for the full process.