Skip to content

Skills Overview

The BigQuery ETL Skills plugin includes specialized skills that work together to accelerate your workflows.

Anatomy of a Skill

Each skill is a folder containing instructions, templates, and reference materials that Claude loads on-demand:

skill-name/
├── SKILL.md                    # Main skill instructions
├── assets/                     # Templates and examples
│   ├── template1.sql
│   ├── example1.py
│   └── ...
├── references/                 # Detailed reference docs
│   ├── conventions.md
│   ├── patterns.md
│   └── ...
└── scripts/                    # Helper scripts (optional)

SKILL.md

The main skill file with three key parts:

1. Frontmatter Metadata 

---
name: skill-name
description: When Claude should invoke this skill
---

The description field is critical—it tells Claude when to automatically invoke the skill based on user requests.

2. Instructions

Step-by-step guidance for completing the skill's task:

  • What to read first (references, templates)
  • How to approach the work
  • What to generate or update
  • Required validations
  • When to invoke other skills

3. Coordination Points

Explicit instructions for working with other skills: 

## Skill Coordination
- After completing X, invoke metadata-manager
- If Y is needed, invoke sql-test-generator

assets/

Templates and examples that Claude can read and copy:

  • SQL query templates
  • Python script templates
  • Configuration examples
  • Sample input/output data

These files provide concrete patterns to follow, reducing hallucination and ensuring consistency.

references/

Detailed documentation broken into focused topics:

  • Conventions and standards
  • Common patterns and anti-patterns
  • External documentation links
  • Decision trees and workflows

Breaking knowledge into smaller files helps Claude load only what's relevant for each task.

scripts/ (optional)

Helper scripts for validation, generation, or analysis—used when bash operations are needed during skill execution.