model-deployment

Deploy ML models with FastAPI, Docker, Kubernetes. Use for serving predictions, containerization, monitoring, drift detection, or encountering latency issues, health check failures, version conflicts.

1.75x

Quality

92%

Does it follow best practices?

Impact

100%

1.75x

Average score across 3 eval scenarios

Securityby

Passed

No known issues

Evaluation results

100%

47%

Production API for a Fraud Detection Model

FastAPI ML server setup

Criteria

Without context

With context

Uses FastAPI

100%

Uses joblib

100%

Pydantic request model

100%

Finite value validator

50%

100%

Liveness endpoint

50%

100%

Readiness endpoint

50%

100%

Batch prediction endpoint

100%

model_version in response

100%

latency_ms in response

100%

Request logging middleware

100%

X-Process-Time header

100%

X-Model-Version header

100%

26%

Containerize a Sentiment Analysis API for Production

Docker containerization

Criteria

Without context

With context

Base image python:3.11-slim

100%

Multi-stage build

100%

Non-root user

100%

MODEL_PATH env var

100%

Uvicorn CMD on port 8000

100%

HEALTHCHECK directive

100%

Pinned requirements

100%

Versioned image tag

100%

Build arg for model version

100%

pip --no-cache-dir

100%

WORKDIR /app

100%

EXPOSE 8000

100%

55%

Deploy a Churn Prediction Service to Kubernetes with Drift Monitoring

Kubernetes deployment and monitoring

Criteria

Without context

With context

Memory request 512Mi

100%

Memory limit 1Gi

100%

CPU request 500m

100%

CPU limit 1000m

100%

Liveness probe /health

100%

Readiness probe /ready

100%

Rolling update maxUnavailable 0

100%

maxSurge: 1

100%

KS test for input drift

100%

Jensen-Shannon for prediction drift

100%

Rolling window deque

100%

should_retrain method

100%

Periodic drift check

100%

Repository: secondsky/claude-skills
Commit: 90d6bd7

Evaluated: 29 days ago
Agent: Claude Code
Model: Claude Sonnet 4.6

Table of Contents

Production API for a Fraud Detection Model Containerize a Sentiment Analysis API for Production Deploy a Churn Prediction Service to Kubernetes with Drift Monitoring

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.