Health: a score that explains itself and a sheet for every task

A score that explains itself

At the top, your repo in a score from 0 to 100, with its band in words and what each dimension takes from it: the tests, the cleanup, the change risk, the security and, if your code uses Sentry, the real errors in production. The most serious thing you have open decides the band, and the sum always adds up.

Since your last visit

What happened to your Health since you last opened it, in one sentence signed by whoever pushed: what broke and got fixed, what's new and whether it's serious. A new error in production comes first, even if nobody pushed.

First, and every task in its sheet

On top goes the task worth fixing before any other: first what happens to your users, then what's exposed and what's broken. Under it, the rest by how serious it is, all in sight, with the errors of one page together. Tap a task and its sheet opens: what it is, why it matters and when it came; what SeeRepo found and where it lives on your map; and the prompt for your agent, ready to copy. Every sheet has its link, and the bell's notice of a serious task takes you to it.

What you fixed, for real

What you fixed in the last month, said as it is now ("Your tests pass again") and with who fixed it. Only what a push closed, or what you marked resolved in Sentry, counts: what closed because the week passed isn't written down as a fix.

Share it, and let your agent read it

Your score in a square image from Share, the Health folder with its ring in the project, and the same Health for your agents in the API, the MCP (health) and the CLI (seerepo health), which also tells them whether it can ship.