wan-huiyan

Dbt Model Planner — Data skill for Claude Code

Data community

Claude Code skill: Plan and build dbt models from schema requirements through structured planning with human review, schema validation, and multi-dialect SQL generation.

How to install Dbt Model Planner

This entry records only its repository, not the path inside it, so there is no exact command to give. Open wan-huiyan/dbt-model-planner and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Dbt Model Planner does

Claude Code skill: Plan and build dbt models from schema requirements through structured planning with human review, schema validation, and multi-dialect SQL generation.

Alternatives in Data

  • Private GPT — Complete API layer for private AI applications on local models: RAG, skills, tools, MCP, text-to-sql, and more 57.5k ★
  • Finance Skills Pack — Financial analysis and planning skills 5.3k ★
  • MCP Server Chart — 🤖 A visualization mcp & skills contains 25+ visual charts using @antvis 4k ★

README

dbt Model Planner

[![GitHub release](https://img.shields.io/github/v/release/wan-huiyan/dbt-model-planner)](https://github.com/wan-huiyan/dbt-model-planner/releases) [![Claude Code](https://img.shields.io/badge/Claude_Code-skill-orange)](https://claude.com/claude-code) [![license](https://img.shields.io/github/license/wan-huiyan/dbt-model-planner)](LICENSE) [![last commit](https://img.shields.io/github/last-commit/wan-huiyan/dbt-model-planner)](https://github.com/wan-huiyan/dbt-model-planner/commits)

Jumping straight from a dashboard brief to dbt SQL skips the hard part: mapping every target field to a source table, validating it exists in production, and agreeing on the logic before writing a line of code. This skill guides you through that planning process with human review checkpoints before anything gets built.

What It Does

Takes a target schema (CSV spec, dashboard brief, or requirements doc) and guides you through 7 phases to build a validated dbt model:

  1. Understand Requirements — parse the spec, determine model grain
  2. Context Gathering — auto-read project config, mine conventions, explore source models, check with stakeholders
  3. Planning Document — structured CSV/Excel mapping every target field to source tables with transformation logic
  4. Human Review Loop — iterate with the team until all fields are confirmed
  5. Schema Validation — DESCRIBE production tables (3-part check: existence + population + freshness), write investigation queries
  6. Build the Model — CTE-structured SQL with recommended conventions and a dialect reference table
  7. Verify and QA — standalone verification query + self-contained QA queries

Key Features

  • Complexity gate — lightweight path for simple models (<8 fields), full 7-phase path for complex ones
  • Multi-dialect support — dialect reference table for Spark/Databricks, Snowflake, BigQuery, and Postgres/Redshift
  • Human review checkpoints — blank "Review Feedback" column in