Skill UI

This skill provides comprehensive guidance for building production-grade MLOps pipelines. It covers the entire machine learning lifecycle, from initial data ingestion and rigorous feature engineering to model training, advanced validation (A/B testing, performance drift), and robust deployment strategies (Canary, Blue-green). It enables the design and orchestration of scalable, reproducible, and observable ML systems using tools like Airflow and Kubeflow.

NeMo Curator streams GPU-accelerated curation for LLM training, with super-fast fuzzy deduplication, multimodal quality filters, semantic dedupl, PII redaction, NSFW checks, and RAPIDS-scaled pipelines for prepping massive corpora.

Automates Neon serverless Postgres workflows via the Composio MCP, letting you list projects and branches, inspect databases, manage roles, and retrieve connection URIs for pipelines and tooling integrations.

This comprehensive reference architecture guides the integration of OpenEvidence into clinical decision support systems (CDS). It is designed for healthcare platforms needing evidence-based query processing, deterministic citation tracking, and full audit logging. Key considerations include HIPAA compliance, optimizing for sub-second response times in point-of-care workflows, and establishing robust data models and event pipelines.

This skill enables the development of robust, event-driven pipelines interacting with the Perplexity Sonar API. Since Perplexity lacks native webhooks, this system simulates event patterns using advanced techniques: Streaming Server-Sent Events (SSE) for real-time data, job queues (like BullMQ/SQS) for asynchronous batch research, and cron-triggered monitoring for scheduled news and trend alerts. It provides patterns for building complex, reliable, and scalable AI applications.

Phoenix is an open-source observability platform for tracing, evaluating, and monitoring LLM applications; use it to debug prompts, run dataset evaluations, build experiment pipelines, and watch production inference in real time.

Guides users through setting up monitoring for data pipelines, covering ETL, transformation, orchestration, streaming, and validation with best-practice code and configuration.

Set up comprehensive CI/CD pipelines for PostHog integrations using GitHub Actions. This guide covers mocked unit tests (no API key needed), robust integration tests against a development PostHog project, and automated deployment annotations to mark releases in your PostHog timeline. Ideal for developers ensuring data quality and traceability across all environments.

This guide provides comprehensive strategies for implementing privacy-safe analytics using PostHog. It addresses critical compliance requirements like GDPR and CCPA, covering explicit consent management (opt-in/opt-out), real-time property sanitization (stripping PII like IP or tokens), and methods for handling Data Subject Access Requests (SAR) and the Right to Erasure. Use this to build highly compliant and privacy-preserving analytics pipelines.

Monitor the operational health of PostHog integrations by tracking critical metrics. This skill checks event ingestion rates, feature flag evaluation latency, and monitors billing volume consumption. It also provides rules for Prometheus/Grafana alerting, ensuring data pipelines remain stable and alerting on potential performance bottlenecks or exceeding API limits.

Guides and automates Prefect Flow Builder tasks for data pipelines, offering best-practice patterns, production-ready code, configuration templates, and validation advice whenever you mention the tool or request pipeline help.

This skill builds and executes automated pipelines to clean, validate, and transform data for machine learning, covering ETL workflows, error handling, and quality reporting so Claude can deliver analysis-ready datasets.