customer-health-check
Skill1 starsChanged 5 months agoMITView on GitHub
Flag at-risk accounts in your customer base and explain why, sorted into Red/Yellow/Green with a concrete next action for each. Use when the user says "churn risk", "at-risk customers", "customer health", "which accounts might leave", "account review", or "find customers about to churn".
In this repository3 other files it gives its agents.
- complaint-de-escalationSkill
- review-responderSkill
- support-reply-drafterSkill
The file
Files
Customer Health Check
Flag accounts that look at-risk and explain why.
When to use
You want to catch churn before it happens — scan your customer base, sort accounts by risk, and get a specific action for each one that's slipping.
What you'll need
A list of customers with recent activity (last order/login date, spend trend, support tickets, plan, etc.).
Instructions
Collect any missing inputs from the user, then run this prompt:
You are an account manager reviewing my customer base to catch churn before it happens.
Here is my customer data (one row per customer):
[paste a table or list — include for each: name, how long they've been a customer, last purchase or login date, spend this period vs. last period, number of open or recent support tickets, plan/tier, and anything else relevant]
Today's date is [date].
For each customer, assess churn risk and sort them into Red (likely to leave), Yellow (watch closely), and Green (healthy). For every Red and Yellow account, give me:
- The 1-2 specific signals that triggered the rating (e.g. "no login in 47 days; spend down 60%").
- A plain-English reason this matters.
- One concrete action I should take this week (e.g. "personal check-in call," "send a win-back offer").
Then list the top 5 accounts to contact first, highest risk and highest value at the top. Don't invent data — if a signal is missing for someone, say what you'd need to judge them.Tip: Include both engagement (logins, usage) and money (spend trend) columns. A customer who's still paying but stopped logging in is often the most dangerous — quietly disengaged and one renewal away from leaving.
Built by Ootto — the AI autopilot that connects your tools once and runs invoicing, follow-up, and reports for you, automatically. Book a demo →
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

