See your whole book in one rollup.
Scores every account on health, lays out renewals due, shows product adoption, and flags accounts going quiet.
Inside: The health composite, the renewals-due view, and the adoption heatmap.
Install it in one line, or paste it in.
~/.claude/skills/ and runs automatically when it is relevant.Connect your context. Set it to your motion.
the whole book with owners, value, and renewal dates, live.
per-account usage and active users, which power the health composite and the going-dark detector.
Map your portfolio first. Tell the skill which owner field equals "my book," which contract-value field is the source of truth for ARR, and what your product line names are. The method runs against those placeholders.
| Set this | What it is | Default / Example |
|---|---|---|
| CRM | Where accounts, owners, and contract values live | |
| Renewal date field | The field the renewal queue reads | A "renewal end date" field |
| Product lines | The set of products you check penetration against | CoreAdd-on AAdd-on BAdd-on C |
| Health-score inputs | The signals that combine into the composite | Engagement tieractive userschampion stabilitysupport sentiment |
| Dark-days threshold | Trailing window of no activity that flags an account | 30 days |
Everything the skill does, in full.
This skill turns your customer book into one rollup view. It pulls every account you own, scores each on a health composite, lays out the renewals coming due, shows which products each account has adopted and which it has not, flags accounts going quiet, and ranks the best expansion plays. It is built for the start of the week, when you need to know where to spend your time. It is a rollup, not a per-deal deep dive.
- 11. Pull the book. Every account where the owner field matches the portfolio owner, with contract value, renewal date, and product lines.
- Score the health composite per account. Combine the inputs you mapped into one of four bands: HEALTHY, WATCH, AT-RISK, GHOST. Every band carries the reasons that produced it.
- Build the renewal queue. Bucket open renewals into 90, 120, 180-day horizons, each with a verdict.
- Build the multi-product penetration view. Per product line, count adopted vs total. The largest gap is the biggest opportunity.
- Run the going-dark detector. Flag accounts past the dark-days threshold or with falling active users. Pair each with contract value and renewal date.
- Compute retention contribution per account. Year-over-year value delta net of churn and downgrades, rolled to a portfolio figure.
- Rank expansion candidates. Single-product accounts at an adoption ceiling carrying the signals to take the next product.
- Close with named next moves tied to specific accounts.
- The health composite is sourced. Never "Acme is at risk," always "at risk because engagement dormant, zero activity in 30 days, champion left in March."
- The renewal queue carries a risk overlay (HEALTHY / WATCH / SAVE_NEEDED).
- Book comparisons are honest. Surface book mix, not just headline numbers. One owner may hold the harder accounts.
- If a query needs a signal you have not mapped, the skill stops and says what to connect.
PORTFOLIO ROLLUP · 47 accounts · $4.2M ARR Health composite: HEALTHY: 31 ($2.8M) WATCH: 9 ($820K) AT-RISK: 5 ($412K) GHOST: 2 ($168K) Renewal pipeline: Next 90d: 6 renewals · $678K · 4 HEALTHY · 1 WATCH · 1 SAVE_NEEDED Next 120d: 11 renewals · $1.2M Multi-product penetration: Core: 42 of 47 (89%) Add-on C: 8 of 47 (17%) (34 expansion candidates, biggest gap) Going-dark this week: - Account One (-87% active users, $42K ARR, renewing 2026-08) Retention contribution YTD: +$340K net · 108% net retention Top expansion candidates: 1. Account Four ($187K), Core only, ready for Add-on C
The defaults are illustrative. Re-tune to your book: the four health bands and their inputs, the 90/120/180-day renewal horizons, the 30-day dark threshold, the product lines and what counts as adopted, and the retention math (the year-over-year delta and what you net out). Match each to your own retention reality.
Where an operator takes this next.
A weekly rollup by hand catches this week's fires. Here's the version that never lets one slip.
The rollup is the read. The system is what makes sure nobody has to remember to ask for it.
Connect Salesforce and your product-analytics tool to a scheduled Claude task that regenerates the rollup before your week starts, no manual export required.
Wire Amplitude usage drops into a Slack DM to the account owner the day the threshold trips, not the week you happen to run the report.
Feed the ranked list into Salesforce as a task or a Clay enrichment trigger so a single-product account at the adoption ceiling becomes a warm handoff, not a line item nobody actions.
One skill is the on-ramp.
A single skill does one job. Chained into a playbook, or run as a full build, it becomes a system. Here is where this one plugs in.