fabriciq

Answers natural-language business questions over Power BI reports via the FabricIQ MCP endpoint.

Updated Jul 14, 2026
One-click install
npx skills add https://github.com/9vantage/skills-for-fabric-clone --skill fabriciq-9vantage
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: fabriciq
Source: https://github.com/9vantage/skills-for-fabric-clone/tree/main/plugins/fabric-skills/skills/fabriciq
Command: npx skills add https://github.com/9vantage/skills-for-fabric-clone --skill fabriciq-9vantage

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Business users often need quick answers from Power BI reports and dashboards without writing DAX queries or navigating complex report structures. This Skill translates natural-language questions into accurate DAX queries against Power BI semantic models and returns plain-language answers. ## Core Features & Use Cases - Artifact Discovery and Resolution: Finds Power BI reports and semantic models by free-text search, and resolves report GUIDs from pasted Power BI or Fabric URLs. - Schema-Aware DAX Generation: Inspects report metadata and semantic model schemas, honors Custom Instructions and Verified Answers, and resolves entity values before writing filters. - Query Execution and Synthesis: Executes up to four DAX queries per call, applies report/page/visual filters, and synthesizes results into concise answers with bolded key numbers. - Use Case: Ask "What are the top 5 products by revenue in the Sales Benchmark report?" and receive a formatted answer derived from the actual report data, with all report filters respected. ## Quick Start Ask the assistant: "What were the top selling products last quarter in my Sales Overview Power BI report?"

Frequently Asked Questions about fabriciq

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I ask natural-language questions about my Power BI report data?

Provide a question plus a report name, keyword, or Power BI URL. The skill discovers the artifact, reads its metadata and semantic model schema, generates a DAX query, executes it, and returns a plain-language answer with the key figures.

What is the difference between FabricIQ and raw DAX execution against a semantic model?

FabricIQ is designed for business questions about report content and orchestrates discovery, schema inspection, and query generation end to end. For raw DAX EVALUATE statements or model metadata inspection with INFO functions, use the semantic-model-consumption skill instead.

Can I use a Power BI URL instead of a report name?

Yes. When you paste a Power BI or Fabric URL, the skill calls ResolveReportIdFromUrl to map it to the correct published report GUID, which is required for workspace-App URLs where the path-level ID is a per-app instance ID.

Which Power BI artifact types are supported for question answering?

Only Power BI reports and semantic models are supported. Paginated reports, dashboards, and other Fabric artifact types are not currently supported, and the skill will suggest using a report or semantic model instead.

Why did my Power BI query return blank or unexpected results?

Blank results usually come from querying dates with no data, missing report-level filters, or using the wrong measure. The skill re-checks schema, filters, and measure bindings, corrects the DAX, and retries once before reporting the issue.

What are verified answers in a Power BI semantic model?

Verified answers are model-owner-authored visual definitions that act as the highest-priority source of truth. When one matches your question, the skill replicates its exact bindings, filters, and granularity in DAX instead of writing a custom query.