pingouin-statlisted
Install: claude install-skill Exekiel179/pingouin-psych-stats
# Pingouin Stat
This is the main entry point. Use it instead of asking users to choose individual sub-skills.
## Load
Read in order:
1. `../../references/intake-checklist.md`
2. `../../references/workflow-index.md`
3. `../../references/supervision-gates.md`
Read `../../references/archive-contract.md` before generating code or a deliverable that must be reproducible.
Read `../../references/workflow-contract.md` for a resumable multi-command run.
Read `../../references/pingouin-optimization.md` before generating concrete Pingouin code.
Read `../../references/pingouin-api-quickref.md` only when generating code. Read `../../references/apa-output-template.md` only when writing final prose or tables.
## Workflow
1. Intake: clarify only missing items that affect the analysis.
2. Route: choose one or more downstream skills from `workflow-index.md`.
3. Screen: use `pg-data-screening` when data shape, missingness, or assumptions are unknown.
4. Analyze: call the selected analysis workflow or generate the relevant Pingouin code.
5. Approve: run `pg-analysis-approval` before final interpretation.
6. Organize: use `pg-reporting` to produce the requested deliverable shape.
7. Archive: initialize a run under `archive/analysis-runs/`, save the exact code and numerical outputs, then write the report and S0–S5 audit record.
## Intake Contract
If key information is missing, ask a compact question set covering:
- Research hypothesis.
- Data path or available columns.
- Outcome, predic