Find agent skills by outcome

description: A Claude Code agent skill for Unity game developers. Provides expert-level architectural planning

system design

Unity Architecture Specialist

---

| MAR

conditioned on group | Group-wise mean/mode |

Coding & Debugging

Kickstart Prompt for UX & UI Design

You're an award winning UX & UI designer who is expert on nextjs, react, tailwind.

| Numerical

time-series | Forward fill / Interpolation |

| Categorical

low cardinality | Mode imputation |

| Categorical

high cardinality | Fill with 'Unknown' category |

→ Check for disguised missing values (?

N/A

| Numerical

symmetric dist | Mean imputation |

| Numerical

skewed dist | Median imputation |

After completing Phases 1–4

deliver this exact report:

[Column] — Reason: [e.g.

72% missing

- Any disguised nulls found (?

N/A

# imp_iter = IterativeImputer(max_iter=10

random_state=42)

assert remaining_train.sum() == 0

fTrain still has missing:\n{remaining_train[remaining_train > 0]}

assert remaining_test.sum() == 0

fTest still has missing:\n{remaining_test[remaining_test > 0]}

X_test = X_test.drop(columns=drop_cols

errors='ignore')

X_train = X_train.drop(columns=drop_cols

errors='ignore')

# STEP 12 — MICE / IterativeImputer (most powerful

use when needed)

X

X_train

X_test

df.dropna(subset=[TARGET_COL]

axis=0

df.replace(DISGUISED_NULLS

np.nan

DISGUISED_NULLS = [?

N/A