data-analysis-skillslisted

# data-analysis-skills: Industry-Aware Data Analysis & Narrative Reporting ## Overview Turn raw data into industry-aware data stories that speak for themselves — the deliverable is not isolated charts and numbers, but an analysis report with insights, conclusions, and the power to drive business decisions. ``` One analysis = five-phase workflow (the main line) × one industry template industries/ (what to analyze, who to speak to) × method modules references/ loaded on demand (how to analyze) → data JSON injected into assets/report_template.html → single-file HTML report ``` Supported inputs: CSV, Excel (.xlsx/.xls), JSON, TSV. A single input may carry **multiple datasets** — an Excel file with several sheets, two or more files handed over at once, or a JSON whose top level is a dict with multiple array keys; handle these via the Phase 1 multi-source intake (see `references/data-sources.md`). Analysis runs on Python (pandas etc.; if a dependency is missing, try pip install first, and fall back per the degradation table on failure). ## First: Task Triage | What the user is asking for | Path | |------|------| | Open-ended analysis, wants a report/insights ("analyze this data", "give me a report") | **Full five-phase workflow** (including the Phase 1 gate), see below | | A single statistical question ("test whether B is significant", "compute the correlation between these two columns") | **Lightweight path**: run that computation direc