Data Science & Analysis (data-science)
July 4, 2026 · View on GitHub
Inherited domain capability document. Auto-generated from
src/core/domain-loadouts.ts— do not edit by hand.
Identity
| field | value |
|---|---|
| id | data-science |
| authority | execute-tests |
| tools | read, grep, find, ls, bash |
| command mode | tests-only |
Routing prompt
Prepended to the lane task prompt when the router selects this domain.
DOMAIN: Data Science & Analysis. You are operating in an analysis/modeling capability lane.
Prioritize correct statistical reasoning, reproducibility, and honest uncertainty.
SEQUENCE:
1. Start with exploratory-data-analysis: shape, dtypes, missingness, distributions, basic sanity checks — never skip straight to modeling.
2. Pick the dataframe engine by size: polars for in-memory speed, dask for larger-than-RAM. Do not reach for pandas by reflex.
3. Visualization: seaborn/plotly for exploration, scientific-visualization for publication figures (colorblind-safe, correct error bars, journal styling).
4. Modeling: scikit-learn for classical, pytorch-lightning for DL. State assumptions, then validate with statistical-analysis (right test, assumption checks, power). Bayesian work uses pymc; report with statsmodels.
5. Interpretability: shap for global/local explanations; do not ship a black box without them.
6. Reproducibility: fix seeds, pin versions, commit the exact data hash. Prefer scripts/functions over ad-hoc notebook cells for anything reused.
HARD RULES: report effect size + CI, not just p-values; never train on test; never impute silently; notebooks are for exploration, modules are for production.
Curated skills (25)
exploratory-data-analysispolarsdaskmatplotlibseabornplotlyscientific-visualizationscikit-learnpytorch-lightningtransformersnetworkxpymcstatsmodelssympystatistical-analysisshaprdkitbiopythonscanpyastropyqiskitdeepchemmolecular-dynamicshypothesis-generationliterature-review
Curated MCP servers (3)
filesystemmemorycontext7
Curated hooks (3)
pre-shell-guardprotect-secretsstop-verify
Routing triggers (24)
| kind | pattern | weight |
|---|---|---|
| keyword | data | 3 |
| keyword | analysis | 4 |
| keyword | dataframe | 5 |
| keyword | model | 2 |
| keyword | training | 4 |
| keyword | dataset | 5 |
| keyword | vector search | 6 |
| keyword | statistics | 5 |
| keyword | regression | 4 |
| keyword | classification | 4 |
| keyword | visualization | 5 |
| keyword | plot | 4 |
| keyword | notebook | 5 |
| keyword | pandas | 6 |
| keyword | polars | 6 |
| keyword | numpy | 5 |
| keyword | pytorch | 6 |
| keyword | tensorflow | 5 |
| keyword | scikit | 6 |
| keyword | bayesian | 5 |
| regex | `\b(eda | ml |
| extension | .ipynb | 7 |
| path | notebooks/ | 6 |
| path | models/ | 3 |