Apache Beam Skills

This directory contains skills that help the agent perform specialized tasks in the Apache Beam codebase. For more information, see the Agent Skills Documentation.

Available Skills

Skill	Description
adding-new-metadata	Guide on how to add and propagate new metadata fields in WindowedValue to avoid metadata loss
beam-concepts	Core Beam programming model (PCollections, PTransforms, windowing, triggers)
beam-dofn-modernizer	Rewrite Apache Beam DoFn methods to remove legacy ProcessContext/OnTimerContext
ci-cd	GitHub Actions workflows, debugging CI failures, triggering tests
contributing	PR workflow, issue management, code review, release cycles
gradle-build	Build commands, flags, publishing, troubleshooting
io-connectors	51+ I/O connectors, testing patterns, usage examples
developing-new-io-connectors	A detailed guide on developing new I/O connectors
java-development	Java SDK development, building, testing, project structure
license-compliance	Apache 2.0 license headers for all new files
python-development	Python SDK environment setup, testing, building pipelines
runners	Direct, Dataflow, Flink, Spark runner configuration
yaml-development	YAML SDK development, environment setup, testing, and key concepts

How Skills Work

Discovery: The agent scans skill descriptions to find relevant ones
Activation: When a skill matches the task, the agent reads the full SKILL.md
Execution: The agent follows the skill's instructions

Skill Structure

Each skill folder contains:

SKILL.md - Main instruction file with YAML frontmatter

---
name: skill-name
description: Concise description for when to use this skill
---
# Skill Content
Detailed instructions...

Adding New Skills

Create a new folder under .agent/skills/
Add a SKILL.md with YAML frontmatter (name, description)
Write clear, actionable instructions in the markdown body