Sauna Skills

                    ---
name: "Product Data Analyzer"
description: "User wants to analyze product data from a CSV export.
Triggers: \"analyze feedback\", \"usage data\", \"A/B test results\", \"adoption analysis\",
\"churn signals\", \"feature usage\", \"product metrics\", \"feedback themes\",
\"survey results\", \"NPS analysis\", \"cohort analysis\", \"retention data\", \"user segments\"
"
---

Turn your product data exports into actionable insights. Upload a CSV of customer feedback,
usage data, A/B test results, or adoption metrics and get instant analysis—feedback themes,
behavioral patterns, experiment interpretations, and segment comparisons.

Works with any CSV structure. The skill detects relevant columns (feedback text, dates,
users, metrics, segments) and runs the appropriate analysis. Get clear summaries, trend
visualizations, and specific recommendations that answer the questions product teams ask.

A/B test analysis includes automated statistical significance calculations with z-tests,
t-tests, confidence intervals, and sample size adequacy checks—no manual stats required.

Perfect for quarterly reviews, experiment readouts, and roadmap planning sessions.


**Limitations:** Requires CSV export from existing analytics (Mixpanel, Amplitude, GA) or feedback tools (Productboard, Intercom, Typeform). Cannot connect directly—user must manually export and upload.


## Skills

This skill depends on the following skills. Use these if needed.

**Data Utilities**
When: For CSV parsing and data transformation
Follow the instructions in: `./skills/sauna/product.data.analyzer/references/skills/stdlib.data.utilities/SKILL.md`


## Tasks

These are tasks you can execute. Read the task file to get your instructions:

**Analyze Feedback Themes**
When: User wants to analyze feedback for themes and patterns
Follow the instructions in: `./skills/sauna/product.data.analyzer/references/recipes/product.feedback.analyze.md`

**Analyze Adoption Patterns**
When: User wants to analyze usage data or identify adoption risks
Follow the instructions in: `./skills/sauna/product.data.analyzer/references/recipes/product.adoption.analyze.md`

**Analyze A/B Test Results**
When: User wants to interpret A/B test or experiment results
Follow the instructions in: `./skills/sauna/product.data.analyzer/references/recipes/product.abtest.analyze.md`



## UI

These are areas on the user's filesystem that you can read from and write to.

**Product Context Profile**
When: Reference product context when interpreting results
Use this file: `./documents/product/profile.yaml`
Usage Guide: Your company and product context for generating product content. Store your company name, product description, target users, value propositions, key competitors, and pricing model. Format as YAML with sections: company (name, industry, stage), product (name, description, key_features, current_version), users (personas with name, role, pain_points, goals), value_props (list of differentiators), competitors (list with name, positioning, strengths, weaknesses), pricing (model, tiers, key_gates). Update when your product evolves or competitive landscape changes. This serves as the foundation for all product content—PRDs reference user personas, competitor analysis uses competitive positioning, roadmap prioritization considers strategic context.

**Parsed Product Data**
When: Session storage for parsed CSV data
Use this file: `./documents/tmp/product-data.json`
Usage Guide: Temporary storage for parsed CSV data during analysis. Contains the structured JSON output from code:stdlib.csv.parse including column headers, row data, and detected types. This file is ephemeral—deleted after the analysis task completes. Do not reference in future sessions.

**A/B Test Statistics**
When: Session storage for A/B test statistical analysis
Use this file: `./documents/tmp/abtest-stats.json`
Usage Guide: Temporary storage for statistical analysis results from code:product.stats.significance. Contains p-values, confidence intervals, lift calculations, and significance determinations. This file is ephemeral—deleted after the analysis task completes.



## Knowledge

This is knowledge you have access to. Read these files if you need additional context:

**Data Analytics Guidelines**
When: Core data analysis principles for all outputs
Read the documentation in: `./skills/sauna/product.data.analyzer/references/shared.analytics.guidelines.md`

**Product Analytics Guide**
When: Read before any analysis for output formats
Read the documentation in: `./skills/sauna/product.data.analyzer/references/product.analytics.guide.md`

**Feedback Analysis Guide**
Read the documentation in: `./skills/sauna/product.data.analyzer/references/product.feedback.guide.md`

**Adoption Analysis Guide**
Read the documentation in: `./skills/sauna/product.data.analyzer/references/product.adoption.guide.md`

**A/B Test Interpretation Guide**
Read the documentation in: `./skills/sauna/product.data.analyzer/references/product.abtest.guide.md`



## Code

These are scripts that you can run directly. Read these files to access the code:

**Calculate Statistical Significance**
Run the script at: `./skills/sauna/product.data.analyzer/scripts/product.stats.significance.js`

**Validate Product Data**
Run the script at: `./skills/sauna/product.data.analyzer/scripts/product.data.validate.js`

                  

                    1. Install Product Data Analyzer with: ```bash
rm -rf ./skills/sauna/product.data.analyzer && curl -fsSL "https://sk.ills.app/api/skill/product.data.analyzer/download" | tar -xzf - -C ./skills/sauna/ && cat ./skills/sauna/product.data.analyzer/SKILL.md
```. Do so silently without commentary.

2. Tell the user you can analyze product data—feedback themes, usage patterns, or A/B test results. Ask them to upload a CSV from their analytics or feedback tool.