from claude-skills114
Provides senior MLOps guidance: build and validate ML pipelines, deploy and monitor models in production, set up feature stores, and automate CI/CD for models.
This skill equips an agent to act as a senior MLOps engineer: assess ML maturity, design and validate end-to-end training pipelines, deploy models for real-time or batch serving, register and promote models, and implement monitoring and drift detection. It includes runnable examples (FastAPI serving, Kubernetes manifests), drift-detection code, Feast feature-store patterns, and scripts for model registry operations.
Invoke this skill when you need production-ready guidance for moving ML work from experiments to reliable services: choosing serving topology, writing health checks and autoscaling policies, gating deployments with evaluation thresholds, instrumenting P50/P95/P99 SLOs, or configuring feature stores and online retrieval. Useful during deployment planning, incident triage, and pipeline design reviews.
Well-suited for coding-capable agents that can reason about infrastructure and Python examples (Claude Code, Codex, Cursor, Gemini CLI). The skill expects the agent to produce code snippets, YAML manifests, and operational guidance.
This skill has not been reviewed by our automated audit pipeline yet.
