stevencarpenter
UserGeneral-purpose AI agent definitions, reusable skills, and registry tooling
Categories
Indexed Skills (17)
data-engineering-guidelines
Use when designing, building, or reviewing data-engineering work — lakehouse table design (Delta/Iceberg, medallion, partitioning, compaction, schema evolution, MERGE/CDC), realtime and streaming pipelines (Kafka, Spark Structured Streaming, Flink, watermarks, exactly-once), data governance, AI/ML data support (feature stores, vector/RAG), and data quality, reliability, and cost.
diagramming-guidelines
Use when a document, slide, design, or review needs a diagram — architecture, sequence, flow, ERD, or state. Figma FigJam is the default and preferred tool; this rubric covers the FigJam-via-Figma-MCP workflow, the board-plus-snapshot deliverable, diagram-type selection, and the narrow fallbacks. Shared across technical-writer, slide-designer, data-engineer, and any agent that draws.
java-guidelines
Use when writing or reviewing idiomatic Java — immutability, records, Optional, generics, error handling, and concurrency.
jj-guidelines
Use when working with Jujutsu (jj) version control — especially git-colocated repos — covering the working-copy-as-commit model, bookmarks, the operation log for undo/recovery, git-colocation safety, first-class conflicts, and the git-habit→jj command map.
python-guidelines
Use when writing, reviewing, or designing Python code where typing, error handling, resource safety, subprocess use, or test design matter.
rust-guidelines
Use when writing, reviewing, or designing Rust code where idiomatic APIs, ownership, async, unsafe, performance, tests, or public crate ergonomics matter.
scala-guidelines
Use when writing or reviewing idiomatic Scala — functional design, immutability, ADTs, type classes, effect systems, and collection performance.
spark-guidelines
Use when designing, building, or reviewing Apache Spark 4 data pipelines — batch and streaming DataFrame transforms, lakehouse sinks (Delta/Iceberg), Spark Connect, Structured Streaming (watermarks, checkpoints, transformWithState, state data source), partitioning/AQE/shuffle optimization, and idempotent MERGE/CDC patterns. Reach for this whenever the user mentions Spark, PySpark, Structured Streaming, lakehouse ETL, or distributed dataframe pipelines even if they do not name Spark 4 explicitly.
spark-pyspark-guidelines
Use when writing or reviewing idiomatic PySpark 4 code — typed DataFrames, pandas UDFs/Arrow, transformWithStateInPandas StatefulProcessor, StructType schemas, Pandas API on Spark, pytest fixtures, and Python packaging for Spark jobs. Trigger for PySpark, Python Spark ETL, Databricks notebooks in Python, or spark-submit Python pipelines even when "Spark" is not spelled out.
spark-scala-guidelines
Use when writing or reviewing idiomatic Scala Spark 4 code — Dataset API and Encoders, case-class schemas, transformWithState StatefulProcessor, typed aggregations, Catalyst-friendly Column transforms, and sbt/Maven Spark module layout. Trigger whenever Scala Spark, Spark SQL in Scala, or Dataset-based pipelines are in scope even if the user only says "Scala ETL on Spark."
sql-guidelines
Use when writing, reviewing, or optimizing SQL — query correctness, schema and migration design, index choices, or SQLite-specific patterns including FTS5.
swift-guidelines
Use when writing or reviewing idiomatic Swift — value types, optionals, protocol-oriented design, error handling, and structured concurrency.
technical-writing-guidelines
Use when writing or reviewing any technical document — choosing the right Diátaxis mode (tutorial, how-to, reference, explanation), keeping the modes from bleeding into each other, and meeting the quality bar for the mode you are in. Shared rubric for documentation-writer, technical-writer, and any agent producing prose for people to read.
terraform-guidelines
Use when writing or reviewing Terraform/OpenTofu — module structure, state, provider/version pinning, variable validation, security, and plan hygiene.
typescript-guidelines
Use when writing, reviewing, or designing TypeScript or JavaScript where strict typing, async correctness, null safety, or runtime-target concerns matter.
kotlin-guidelines
Use when writing or reviewing idiomatic Kotlin — null safety, immutability, data/sealed classes, coroutines, and scope functions.
tool-priority
Use when choosing between skills, MCP/LSP/code-intelligence tools, specialized agents, and built-in file or shell tools.
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.