book-of-business-triagelisted

# Book of Business Triage You are running the Monday plan for someone who owns more accounts than they can serve. The output is not a list of at-risk accounts — `churn-risk` already produces that. It is a **work queue**: named accounts, named actions, hour estimates, and a running total that stops at the edge of a real capacity budget. Anything that does not fit is written down as a deliberate omission with a date. The rookie version sorts by health score and works down the list until Friday. It plans against 40 hours that do not exist, spends nine of them on a Critical $12k account while a Watch-band $400k account passes its opt-out date in 26 days, and never writes down what it skipped — so an account goes 90 days untouched and nobody can point at the moment the decision was made, because none was. The elite version states the hours honestly, reserves a weekly touch for the top ARR decile whatever its health band, ranks the rest by **dollars protected per CSM hour**, and treats the not-this-week list as a first-class artifact with a deferral counter. Read `../cs-context/references/evidence-standard.md` first. Every account row carries provenance, every derived figure shows its arithmetic, and the queue ends with a Coverage Ledger and a confidence cap — a book with no interaction data, ranked purely on ARR, is a guess in a spreadsheet. ## Before Starting 1. **Read `.agents/cs-context.md`.** Without segment boundaries in dollars, the notice period and the coverage mod