Today's Agent Skill: Customer Survey Insight Summarizer
What It Does
Most small businesses collect customer survey data and then let it sit in a spreadsheet because turning 200 open-ended responses into actionable takeaways takes hours they do not have. This skill does that analysis in minutes and hands you a report with specific things to fix.
How It Works
The agent reads your raw survey export — open-ended text responses, numeric ratings, and multiple-choice answers — and produces a structured insight report: the top five themes customers raised, the sentiment breakdown for each, your highest and lowest-rated items, and three concrete action recommendations ranked by how frequently customers mentioned them.
How to Deploy It
Drop a CSV or Excel export from Typeform, Google Forms, SurveyMonkey, or any tool with a text column into Claude Code, Gemini CLI, Cursor, or Windsurf and run this skill. Best used after any round of structured customer feedback — a post-purchase survey, a product launch questionnaire, or your annual NPS survey. The insight report goes directly to the owner or team lead responsible for each area.
SKILL.md — Ready to Deploy
## Description
Analyzes raw customer survey exports and produces a structured insights report: top themes, sentiment breakdown, highest and lowest-rated items, and ranked action recommendations.
## Trigger
Invoke when the user provides a survey CSV and says "analyze my survey results," "summarize customer feedback," "what are customers saying," or "pull insights from this survey."
## Input
- Survey response CSV with columns: response_id, date, rating (optional, 1-10 or 1-5), open_text_response
- Survey context: what was asked, how many customers were surveyed, what time period it covers
- Optional: product or service name the survey covers
- Optional: previous summary file for trend comparison
## Steps
1. Count total responses and note the date range.
2. For numeric rating questions, calculate mean, median, and distribution across response buckets.
3. For open-text responses, identify recurring themes by scanning for frequently co-occurring words and concepts. Group responses into 5 named themes.
4. For each theme, count how many responses contain it and assess overall sentiment (positive, neutral, or negative) based on surrounding language.
5. Identify the top 3 specific pain pCopy the full SKILL.md and drop it into your agent's skills directory to activate this skill.