bootstrap-domain

Agentic systems design patterns for autonomous AI agents, multi-agent architectures, tool design, MCP integration, memory systems, and agent safety. Covers agent loop design, ReAct patterns, orchestrator/worker patterns, tool definitions, human-in-the-loop workflows, agent evaluation, LangGraph, CrewAI, AutoGen, Claude agent SDK, and production agent deployment. Use when designing or building agents that make decisions and take actions, multi-agent pipelines, tool-using systems, MCP servers/clients, or autonomous workflows. Use when user mentions "agent", "agentic", "autonomous", "multi-agent", "MCP server", "tool use", "orchestrator", "subagent", "memory system", or "agent loop". Do not use for LLM API integration without agent behavior (use llm-engineering), general software engineering, or analytics work.

Analytics engineering practice patterns for dbt, SQL modeling, data transformation, dimensional modeling, metrics layers, and semantic models. Covers dbt Core, dbt Cloud, Jinja, YAML configuration, testing, documentation, and data modeling best practices. Use when reviewing or building dbt projects, SQL transformations, or data models. Do not use for BI dashboards or visualization (use analytics), data pipeline infrastructure (use data-engineering), or ML model development (use data-science).

Analytics practice patterns for dashboards, reports, metrics, KPIs, and business intelligence. Covers Looker, Tableau, Power BI, Metabase, SQL analysis, metric definitions, and data visualization best practices. Use when reviewing or building dashboards, reports, metrics, or analytical queries. Do not use for dbt/SQL data transformation (use analytics-engineering), data pipeline engineering (use data-engineering), or ML model development (use data-science).

Data engineering practice patterns for data pipelines, ETL/ELT, orchestration, data quality, and infrastructure. Covers Airflow, Dagster, Prefect, Spark, dbt, data lakes, lakehouses, Iceberg, Delta Lake, warehouses (Snowflake, BigQuery, Databricks), streaming (Kafka, Kinesis, Flink), CDC, data contracts, data governance, data mesh, and batch processing. Use when reviewing or building data pipelines, ingestion systems, stream processing, data infrastructure, or data platform architecture. Do not use for SQL modeling or dbt projects (use analytics-engineering), business dashboards (use analytics), or ML model training (use data-science).

Data science and machine learning practice patterns for notebooks, model development, experimentation, feature engineering, MLOps, and business analytics. Covers Jupyter, Python, pandas, scikit-learn, PyTorch, TensorFlow, MLflow, Weights & Biases, feature stores, model serving, A/B testing, statistical significance, causal inference, forecasting, and advanced analytics. Use when reviewing or building ML models, notebooks, experiments, ML infrastructure, A/B tests, or statistical analyses. Do not use for production data pipelines (use data-engineering), dbt/SQL models (use analytics-engineering), or business dashboards (use analytics).