Data Analysis Prompt Framework skill for Claude and Codex
Turns a dataset and a business question into a precise analysis plan: the right methods, the metrics to compute, the charts to draw, and a summary shaped for…
Install it once and Claude or Codex runs it for you. When your request matches, your AI loads the skill, asks for anything it is missing, and works through the template itself. Prefer to fill it in by hand? See the full prompt template.
Download the skill (.zip) · View SKILL.md
When it is used
Use when an analyst or manager has a dataset and a business question and needs a structured analysis with findings tailored to a stakeholder. Do not use for writing SQL or Pandas code or building a live dashboard. Related topics: data, analytics, analysis.
What it will ask you for
- Data type (choose one: Sales data, Customer behaviour, Marketing metrics, Financial data, Survey results, Web analytics, Product usage, HR data, Other)
- Data source (free text, e.g. "Google Analytics or Salesforce or an Excel export")
- Time period covered (free text, e.g. "Last 90 days or Jan to Dec 2024")
- Approximate record count (free text, e.g. "5000 rows or 1M events")
- The question to answer (longer free text)
- Key metrics (longer free text)
- Audience for the results (choose one: Executives, Marketing team, Engineering, Investors, Myself)
- Analysis types (choose any: Descriptive statistics, Trend analysis, Cohort analysis, A/B test evaluation, Funnel analysis, Segmentation, Correlation analysis, Forecasting, Anomaly detection)
- Preferred visuals (choose any: Bar chart, Line chart, Heat map, Table, Scatter plot)
- Report format (choose one: One-page executive summary, Detailed report, Slide deck outline, Dashboard mockup)
How it works
- Restate the decision the analysis informs and the metrics that answer it.
- Check data quality (gaps, duplicates, outliers, definition changes) and record how each issue affects the result.
- Run each chosen analysis type and state the method, the finding, and the confidence in it.
- Match a visual to every finding and say what the reader should notice.
- Write the summary for the stated audience, leading with the answer and ending with caveats and next steps.
What you get
A report in the chosen format: answer first, then data-quality notes, findings per method with visuals, caveats, and recommended next analysis.
What it watches out for
- Reporting correlations as causes when the data cannot support a causal claim.
- Skipping data quality checks, so a tracking change shows up as a trend.
- Writing for the analyst instead of the audience, burying the answer under method detail.
How to install
Claude.ai and the Claude desktop app
- Download the skill .zip.
- Open Settings, then Capabilities, and find Skills.
- Choose Upload skill and select the .zip.
- Ask Claude for help with the task. It loads the skill when your request matches.
Claude Code
- Download the skill .zip.
- Unzip it into ~/.claude/skills/ (or .claude/skills/ inside a project).
- Start a new session. Claude Code loads the skill when your request matches.
Codex
- Download the skill .zip.
- Unzip it into ~/.agents/skills/ (or .agents/skills/ inside a project).
- Codex picks it up on its own. Type $ and the skill name to use it, or just ask for the task.
Cursor
- Download the Cursor rule (.mdc).
- Save it in .cursor/rules/ inside your project.
- Cursor's agent applies the rule when your request matches its description.
AGENTS.md (Codex, Cursor, and other coding agents)
- Copy the AGENTS.md section.
- Paste it into AGENTS.md at the root of your repository, creating the file if needed.
- Coding agents that read AGENTS.md follow it when your request fits.