Render computational documents to markdown (DEFAULT), PDF, HTML, Word, and presentations using Quarto. PREFER markdown output for composability. Use for static reports, multi-format publishing, scientific documents with citations/cross-references, or exporting Jupyter notebooks. Triggers on "render markdown", "render PDF", "publish document", "create presentation", "quarto render", or multi-format publishing needs.
Google Cloud Platform and Vertex AI patterns, quirks, and SDK usage for Claude/Anthropic models on Vertex AI. Use when working with GCP, Vertex AI, the Anthropic Vertex SDK, or deploying Claude models on Google Cloud.
Data pipeline architecture patterns and best practices, including medallion/three-layer architecture (Raw/Staging/Enriched or Bronze/Silver/Gold), YAML-based schema management, and ETL workflow patterns. Use when designing or implementing data pipelines, working with data warehouse layers, or managing table schemas in YAML.
DORA engineering metrics project at ~/src/dora. Load when: querying DORA BigQuery views (deployment frequency, lead time, change failure rate, alerts, review time) from any project; joining against DORA.unified_identity or DORA_clean.* views from any project; running the data pipeline (just refresh, just download-*, just upload-*); making OpenTofu infrastructure changes to DORA tables or views; working with team attribution, team identity, or engineer roster data.
Node.js and JavaScript development patterns, gotchas, and library compatibility notes. Use when working with Node.js projects, TypeScript, React, antd, Zustand, Playwright, Bun, or Vitest.
Delegate research and context-gathering tasks to a sub-agent to protect the primary context window. Use when the user asks to "research X", "look into X", "find out about X", "gather context on X", or any investigative framing where answering requires 2+ searches or multiple sources. Also use proactively before starting substantive work when prior context is unknown. Never run research inline — always delegate.
Guide for creating and iterating on skills — modular packages that extend Claude with specialized workflows, tool integrations, domain expertise, and bundled resources (scripts, references, assets). Use when: (1) creating a new skill from scratch ('create a skill for X', 'build a skill that does Y', 'I need a skill for Z'); (2) updating or improving an existing skill ('add a script to the foo skill', 'the bar skill is missing X', 'iterate on the baz skill'); (3) packaging a skill for distribution ('package the skill', 'generate a .skill file'); (4) understanding skill design patterns ('how should I structure this skill?', 'when should I use a script vs a reference?').
Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK). Covers production patterns: usage tracking, resources vs tools, stdio transport gotcha, tool config gating, and testing without a real client.