agentleFS
Sign inStart free
stevemarksavastevemarksava/superteam/skills/fabric-power-bi/SKILL.md

fabric-power-bi

Skill1 starsChanged 2 months agoMITView on GitHub

Power BI semantic modelling on Fabric — star schema, DAX, deployment pipelines, accessibility, row-level security.

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Never build a semantic model with bidirectional relationships unless there is a documented reason

    line 32
  2. Distinctive:

    Every measure lives in a display folder — a flat measure list is a code smell

    line 33

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

/fabric-power-bi — Semantic Models and Reports on Fabric

Power BI on Fabric consumes gold-tier Delta tables via DirectLake or Import mode. A good semantic model is a star schema, has one DAX pattern per measure family, and ships with row-level security wired to Entra groups. Reports are accessible, not just pretty.

When to invoke

  • A Fabric medallion pipeline has produced gold-tier tables ready for consumption
  • Marlo + Atlas are pairing on data viz for a client-facing dashboard
  • A report needs to be promoted dev → test → prod via a deployment pipeline
  • RLS or accessibility compliance is a stated requirement

Steps

  1. Dataset design. Choose storage mode: DirectLake for large gold tables in OneLake (no import, native Delta read), Import for small dimensions or when transformations are needed, DirectQuery only when freshness beats latency. Document the choice per table.
  2. Semantic model — star schema. Facts in the centre, dimensions on the rim. One fact per subject area. Conforming dimensions across facts. Mark date tables as such. Hide surrogate keys from report view.
  3. DAX measures. Group measures into display folders by subject area. Use variables (VAR ... RETURN ...) for readability. Prefer CALCULATE + explicit filter context to iterator functions where possible. Naming: <Subject> - <Metric> (<Unit>) — e.g., Sales - Revenue (GBP).
  4. Deployment pipeline. Wire dev → test → prod workspaces via Fabric deployment pipelines. Parameterise data-source bindings per stage. Never edit prod directly — every change flows through dev.
  5. Accessibility. Alt text on visuals, high-contrast palette, tab order set per page, no colour-only encoding, screen-reader-friendly titles. Run the built-in Accessibility checker before publishing.
  6. Row-level security. Define RLS roles in the semantic model. Filter DAX per role. Map roles to Entra security groups in Fabric — never map to individual users. Test each role with "View as" before promotion.

Deliverable format

  • .pbip project (Power BI Project format) checked into Git — semantic model as TMDL, report as .Report
  • dax-measures.md — measures grouped by display folder, with formula and description
  • rls.md — role list, DAX filter per role, Entra group mapping
  • Accessibility checker report screenshot on final publish

Non-negotiables

  • Never build a semantic model with bidirectional relationships unless there is a documented reason
  • Every measure lives in a display folder — a flat measure list is a code smell
  • RLS is tested per role before promotion to prod
  • Accessibility checker passes before publish — no exceptions

References

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free