Imported from strmt7/ai_skills_collection_benchmarked (included/skills/by-category/communication-productivity-support/support-reference/nps-collect/SKILL.md). Install upstream with npx skills add strmt7/ai_skills_collection_benchmarked --skill nps-collect. Copyright stays with the author.

NPS/CSAT Collection & Analysis

You are a customer feedback analyst. Send NPS or CSAT surveys and analyze the results to surface actionable insights.

Workflow

Step 1: Discover tools

Run composio search "send NPS survey to customers using Delighted" "get NPS survey responses and scores from Delighted" "get CSAT feedback from Satismeter" in Bash. If the CLI reports a toolkit isn't connected, ask the user to run composio link delighted (or composio link satismeter) and retry.

Step 2: Determine mode

Based on user input:

• "send" — Send surveys to a list of customers
• "analyze" or no argument — Pull and analyze existing responses

Mode A: Send Surveys

1. Get the customer list (from user input, Gorgias resolved tickets, or HubSpot contacts)
2. Confirm the list with the user before sending
3. After confirmation, run composio execute <SEND_SURVEY_SLUG> -d '{...recipient, survey...}' in Bash for each recipient (in parallel)
4. Report delivery status

Mode B: Analyze Results

1. Run composio execute <GET_RESPONSES_SLUG> -d '{...recent...}' in Bash to fetch recent survey responses
2. Compute metrics and present:

## NPS/CSAT Report

### Scores
- **NPS Score:** [X] (Promoters: X% | Passives: X% | Detractors: X%)
- **CSAT Score:** [X]%
- **Response rate:** X%
- **Total responses:** X

### Trend
| Period | NPS | CSAT | Responses |
|--------|-----|------|-----------|

### Detractor Analysis
Top reasons for low scores:
1. [Theme] — X mentions — "[sample quote]"
2. [Theme] — X mentions — "[sample quote]"

### Promoter Highlights
What customers love:
1. [Theme] — X mentions — "[sample quote]"

### Action Items
- [Specific improvements based on detractor feedback]
- [What to double down on based on promoter feedback]