How to use the Customer Success Manager agent skill

What is an agent skill and what does this one do?

An agent skill is a folder of instructions, reference material, templates and scripts that an AI agent can load when a task matches the skill's description. This skill is for customer success work. It is described as monitoring customer health, predicting churn risk and identifying expansion opportunities using weighted scoring models for SaaS customer success.

Rather than guessing, the agent runs three bundled Python scripts. They produce deterministic results, so the same input gives the same output. The skill also includes templates for quarterly business reviews, onboarding checklists, success plans and executive business reviews.

• health_score_calculator.py: rates each account Green (75-100), Yellow (50-74) or Red (0-49).
• churn_risk_analyzer.py: assigns Critical, High, Medium or Low churn risk tiers with suggested intervention levels.
• expansion_opportunity_scorer.py: ranks upsell, cross-sell and expansion opportunities by priority and estimated revenue.

Install the customer-success-manager skill

Download the skill as a ZIP file from the page where you found this guide. Then choose one of the two install routes below.

For a personal install that is available across your projects, unzip the download into ~/.claude/skills/. To limit the skill to one project, unzip it into that project's .claude/skills/ folder. The skill folder should contain SKILL.md along with the assets, references and scripts folders.

If you use the Claude apps instead, upload the ZIP under Settings → Capabilities → Skills.

• Personal install: ~/.claude/skills/
• Project install: .claude/skills/ inside your project folder
• Claude apps: Settings → Capabilities → Skills, then upload the ZIP

Prepare your customer data file

All three scripts read one JSON file passed as a positional argument. The file must follow the schema the scripts expect, and the skill's sample file shows a complete example. Each script needs different nested fields, so check the requirements for the script you plan to run.

The health score calculator needs usage, engagement, support and relationship objects, plus previous period scores for trend analysis. The churn risk analyzer needs usage_decline, engagement_drop, support_issues, relationship_signals and commercial_factors, plus a contract_end_date. The expansion scorer needs contract, product_usage and departments objects.

The source does not give a full field-by-field specification in this guide, so start from assets/sample_customer_data.json and match its structure.

• Required for every customer: customer_id, name, segment and arr.
• Segments used in the scripts: Enterprise, Mid-Market and SMB.
• Missing or wrongly typed fields will cause the script to fail or give weak results.

Run the scoring tools

Run the scripts from the skill folder so the relative paths resolve. The default output is readable text in the terminal. Add --format json when you want machine-readable output for a spreadsheet, a pipeline or a later step.

For a portfolio-wide run, the skill's workflow saves each result to its own file. Check that each output file is non-empty before using it in the next step.

• Output formats: text (default) or json via --format.
• Requires Python 3.7 or newer. Check with python --version.
• Uses only the Python standard library, so there are no packages to install.

Ask the agent for an analysis in plain language

You do not need to name the scripts. The skill is triggered when you talk about the topics in its description, such as analysing customer accounts, reviewing retention metrics, scoring at-risk customers, churn, customer health scores, upsell opportunities, expansion revenue, retention analysis or customer analytics.

Name the file and say what you need. The more specific the request, the more useful the output. The agent can also fill in the QBR, onboarding and success plan templates for a named customer.

• Score the health of every account in customer_portfolio.json and flag the red ones.
• Which accounts are at critical or high churn risk this quarter, and what should I do about each?
• Rank upsell opportunities among my healthy accounts by priority.
• Draft a QBR for Acme using the QBR template and the health score from the latest run.

Read and act on the results

Treat the output as a prioritised starting point, not a final decision. The churn tiers come with suggested intervention levels, which map to the risk bands: Critical means immediate executive escalation, High means urgent CSM intervention, Medium means proactive outreach and Low means standard monitoring.

The weights and thresholds are generic defaults. The reference guides cover the scoring methodology, intervention playbooks and industry benchmarks, which help you judge whether the defaults fit your business.

• Health: Green, Yellow, Red.
• Churn risk: Critical, High, Medium, Low.
• Expansion: ranked upsell, cross-sell and expansion candidates.
• Check the references/health-scoring-framework.md before changing any weights.