What it is
Repository health is a per-repository summary of engineering signal: Open PRs, CI success rate, Security alerts, and an overall Flow health figure. It's keyed to the exact repository connected to your workstream (its full org/repo name), so a workstream backed by two repositories has two repository health records.
How to read it
Open PRs and CI success rate are the same kind of engineering-throughput signal Operation View tracks per iteration, just rolled up at the repository level instead of the sprint level. Security alerts count outstanding flags from your provider's dependency/security scanning. Flow health is a single number that blends these into one at-a-glance figure. If a repository's data hasn't synced recently, its card is marked "Data may be stale" rather than silently showing an old number as current.
What to do about it
Watch CI success rate and security alerts at the repository level even in sprints where Operation View looks fine — a repository can accumulate CI failures or new security alerts across several sprints before either one shows up as an iteration-level problem. Read this alongside Spotting Delivery Bottlenecks with Operation View when you're trying to tell whether a slowdown is specific to one sprint or a standing property of the repository itself.