data-goblin/power-bi-agentic-development
Power BI AI skills and Power BI agents for Claude Code and GitHub Copilot: a plugin marketplace of Power BI skills, subagents, and hooks for semantic models, DAX, TMDL, reports, and AI dashboards. Includes Microsoft Fabric skills and Fabric agents. Weekly updates.
What it solves
This project provides a marketplace of plugins designed to teach AI coding agents (such as Claude Code or GitHub Copilot) how to develop, manage, and optimize Power BI and Microsoft Fabric artifacts. It bridges the gap between general-purpose LLMs and the specialized domain knowledge required for semantic models, reports, and workspace management.
How it works
The project is structured as an Anthropic-format plugin marketplace. It provides three primary types of resources that agents can install and use:
- Skills: Domain-specific knowledge and workflows that activate automatically based on context or can be manually invoked to guide the agent through tasks like writing DAX or M expressions.
- Agents: Autonomous subprocesses that handle complex, multi-step tasks, such as auditing semantic models or reviewing Deneb visual specifications.
- Hooks: Deterministic scripts that run automatically after tool use to validate files (e.g., checking TMDL syntax or DAX references) and catch errors early.
Who it’s for
Data engineers, BI developers, and analysts who use AI coding agents to accelerate their Power BI and Microsoft Fabric development workflows.
Highlights
- Comprehensive Tooling: Includes plugins for Tabular Editor, Power BI Desktop, PBIP project files, and the Fabric CLI.
- Automated Validation: Uses hooks to enforce referential integrity, metadata standards, and schema validation.
- Specialized Reviewers: Dedicated agents for reviewing R, Python, and SVG visuals within reports.
- Broad Compatibility: Works with both Claude Code and GitHub Copilot CLI.