powerbi-report-planning

Power BI Report Planning Skill

This skill orchestrates the full lifecycle for a new Power BI report:

Define -> Inspect -> Spec -> Approve -> Build -> Validate -> Publish

It is intentionally broader than a pure requirements-gathering flow: it captures the report spec and continues into implementation after the user approves.

Must/Prefer/Avoid

MUST

• Use this skill for broad report creation workflows that need requirements, dependency checks, approval, and build sequencing.
• Ask focused clarification questions one at a time and stop after the required decision is clear.
• Lock _brief/report-spec.md and get approval before implementation.
• Route design decisions through powerbi-report-design and file mechanics through powerbi-report-authoring.

PREFER

• Infer obvious answers from the prompt, model, existing PBIP files, or earlier rounds instead of re-asking.
• Inspect the semantic model before finalizing page scope or visual recommendations.
• Treat publishing as optional and only proceed when approved.

AVOID

• Do not use this skill for small, surgical edits to existing PBIR files.
• Do not build before the user approves the locked report spec.
• Do not duplicate detailed visual-design or PBIR-authoring guidance that belongs to companion skills.

Examples of When to Use

Use this skill when the user wants to create a new Power BI report and needs both:

1. A guided requirements workflow.
2. A path to implementation after approval.

Examples:

• "Let's create a new Power BI report from this semantic model."
• "Help me define and then build a Power BI report."
• "Use the report playbook to start a new report."
• "Create a reusable workflow for new Power BI reports."

Do not use this skill for a small edit to an existing report page. For direct PBIR authoring tasks, use powerbi-report-authoring. For visual critique or greenfield design guidance without the full guided workflow (a one-off "redesign this" or "what should this look like?"), use powerbi-report-design directly. This skill uses the design skill during Rounds 3–4 — it does not replace it.

Required Operating Rules

1. Ask one question at a time. Use ask_user for each clarification.
2. Run 3-5 clarification rounds maximum. Each round may have one primary question and, only if absolutely necessary, one follow-up.
3. Inspect the semantic model before locking the spec. Use the semantic model skill or an MCP server when available; otherwise inspect local TMDL/PBIP files directly.
4. Check dependencies explicitly. Do not assume Desktop, MCP, authoring, or Fabric publishing are available.
5. Produce one locked _brief/report-spec.md before building.
6. Ask for approval before implementation. Do not build until the user explicitly approves.
7. When approved, build end-to-end. Model changes, PBIR generation, validation, Desktop preview, screenshot loop, and optional Fabric publish.
8. Local edits stay local unless publishing is approved.
9. Do not re-ask known answers. If the original prompt, inspected files, or a prior round already provides audience, page count, delivery target, scope, or design direction, capture it in working notes and move on. Ask only for genuine ambiguity or risky tradeoffs.

Dependency Checklist

Before implementation, capture this status:

If a dependency is unavailable, continue planning and mark the affected phase as blocked/manual. Do not pretend it is available.

Round Structure

Round 0 — Setup and Dependency Check

Goal: identify the semantic model, report target, and available tooling.

Ask only what cannot be inspected automatically:

What semantic model or dataset should this report use?

Recommended choices should be concrete if candidates are discoverable. Enumerate existing Fabric semantic models and local .SemanticModel folders in scope, then present them as options and mark the best match as recommended. Example:

• <DiscoveredSemanticModelName> (Recommended)
• Another existing Fabric semantic model
• A new local dataset

Then inspect/check:

• Existing .pbip, .Report, .SemanticModel folders.
• Whether TMDL files exist.
• Whether Power BI Desktop automation script exists.
• Whether MCP connection is available or can be established.
• Whether Fabric publishing is requested.

Output working notes:

Dependency status:
- Semantic model:
- PBIP/PBIR:
- Desktop preview:
- Modeling MCP:
- Report-authoring skill:
- Fabric publishing:
- Node generator:

Round 1 — Audience and Job

Goal: understand who the report is for and what decision/job it supports.

If both audience and job-to-be-done are already clear from the prompt, summarize the inferred answer instead of asking. Otherwise ask for the missing piece(s), one at a time.

Ask when audience is unclear:

Who is this report primarily for?

Recommended choices:

1. Executives / leadership — concise KPIs, trends, risks, decisions
2. Analysts — exploration, drilldowns, comparisons, tables
3. Operators — monitoring, exceptions, status, action queues
4. External audience — polished story, guided narrative, minimal slicers
5. Enthusiasts / fans — magazine-style narrative, rich visuals, rankings

Then ask only if the job-to-be-done is still unclear:

What should the report help them do?

Recommended choices:

1. Understand the overall story
2. Track performance
3. Find outliers or opportunities
4. Compare entities or segments
5. Explore individual records/profiles
6. Prepare for a recurring business review

Capture:

Audience:
Primary purpose:
Tone:
Success criteria:

Round 2 — Model Inventory and Scope

Goal: inspect the model and define the first-build scope boundary without re-asking for scope the user already gave.

Use whatever model-inspection capability is available — pick the first that applies, in order of preference:

1. A semantic-model authoring skill, if installed.
2. A modeling MCP server (e.g., powerbi-modeling-mcp-*) — connect to the live model, list tables/columns/measures/relationships, and run DAX for live validation.
3. Direct reads of local TMDL files (definition/model.tmdl, definition/tables/*.tmdl, definition/relationships.tmdl) when no live connection is available.

Do not hardcode tool names in user-facing prompts — describe the inspection intent and let the agent pick the available tool.

Summarize:

Facts:
- table, grain, keys, useful measures

Dimensions:
- date/time, geography, entity, category, owner, status, segment

Existing measures:
- core available measures

Likely missing model work:
- measures
- calculated columns
- relationship fixes
- sort columns
- helper fields for slicers/search

Risks:
- nulls/sparsity
- inactive relationships
- high-cardinality slicers
- fields that will not filter as expected

After inspection, infer the first-build scope from the original request and prior answers. Do not ask a standalone scope question if the user already gave a page count, report type, or content boundary (for example, "build a 4 page dashboard"). Instead, summarize the boundary in working notes:

First-build scope:
- <focused first build / standard report / operational app / narrative report>
- inferred from: <user prompt | prior answer | model shape>
- included now:
- deferred:

Ask a scope question only when the boundary is unclear, too broad, or risky. If asking, present tailored choices from the inspected model rather than generic first-build options. Example:

The model supports sales, products, geography, discounts, and profitability. For this first build, should we keep it focused on executive performance, or include a deeper product/customer exploration too?

Round 3 — Narrative and Page Plan

Goal: turn the model and scope into page architecture.

Invoke or explicitly consult powerbi-report-design for page-level archetype routing and composition guidance. The design skill owns visual routing; do not duplicate its routing table here. Use the data shape from Round 2 to surface 2-3 report shape options for user sign-off.

The five archetypes the design skill ships are: Executive Summary, Operational Monitor, Analytical Canvas, Narrative Story, Comparative Benchmark. Use the design skill's archetype and composition guidance for layout variants, multi-archetype reports, and cross-page variant rotation.

Ask only after applying the inferred first-build scope from Round 2:

Which report shape should we use?

Present 2–3 named compositions (e.g., Executive landing + Analytical exploration + Comparative ranking for a multi-domain ask, or Single executive landing for a focused ask). Recommend one based on Rounds 1–2 and mark it (Recommended).

Draft page list in the answer after the user chooses:

Proposed pages:
1. <Page> — archetype — purpose — core visuals — fields/measures
2. <Page> — archetype — purpose — core visuals — fields/measures
<repeat for each approved page>

Capture slicers and interactions:

• Global slicers
• Page-specific slicers
• Search/prefix slicers for high-cardinality dimensions
• Drillthrough/profile pages
• Bookmarks/navigation if needed

Round 4 — Design Identity, Accessibility, and Delivery

Goal: lock the design identity, accessibility baseline, and delivery target.

Invoke or explicitly consult powerbi-report-design for design identity and theme direction. Design identity (tone + signature) is owned by the design skill — do not invent a parallel vocabulary here. Use the design skill's identity guidance to pick a tone+signature combination, then surface it to the user.

Ask:

What should this report feel like?

Present 2–3 concrete identity options adapted to the audience and domain from Rounds 1–2. Each option names a report feel and the one signature visual move it implies. Recommend one and mark it (Recommended).

If the user has brand guidelines, make one option brand-forward rather than picking a generic tone.

Then ask delivery only if unclear:

Where should the finished report end up?

Recommended choices:

1. Local PBIP only first
2. Publish to an existing Fabric workspace
3. Create new semantic model + report in Fabric
4. Update an existing Fabric report
5. Build local first, decide publishing later

Design defaults (these come from the design skill's gotchas + base theme; applied automatically unless the user overrides):

• Use accessible contrast (WCAG AA minimum on every text/background pair).
• Avoid red-on-red and low-contrast palettes.
• Prefer Azure Map over deprecated map/filledMap visuals.
• Add alt text to every chart.
• Use searchable dropdown slicers for high-cardinality fields.
• Use tile/list slicers only for short categorical fields.
• Place detailed tables near the bottom of the page.
• Keep report interactions predictable and consistent.

Design Contract Gate

Before producing _brief/report-spec.md for approval, get a canonical Design Brief: YAML block from powerbi-report-design. The planner may provide requirements, model inventory, page goals, and user constraints to the design skill, but the planner must not author a competing detailed design skeleton.

The canonical design block must include:

• generated_by: powerbi-report-design
• contract_version
• one pages[] entry for every page in the page plan
• pages[].layout_contract.canvas
• pages[].layout_contract.grid.regions
• pages[].layout_contract.placements
• pages[].layout_contract.space_audit
• one page_title textbox placement with non-empty title text per page
• slicer placements in a top-right filters region or a justified filter rail
• no bare single-value cardVisual occupying the largest/dominant hero region unless the design marks it as a composite KPI treatment with context and rationale
• no unresolved placeholders, ellipses, or prose-only wireframes

If the block is missing these items, stop and revise the design contract before asking for approval or invoking powerbi-report-authoring.

Locked Report Spec Output

After Rounds 0-4, produce one file and save it under ./_brief/ in the current working directory:

• ./_brief/report-spec.md — the single source of truth for approval and implementation handoff.

report-spec.md has two layers:

1. Markdown sections for user approval and readable context.
2. A fenced yaml block containing the exact Design Brief: returned by powerbi-report-design — the canonical implementation contract that powerbi-report-authoring consumes.

If Markdown prose and the embedded YAML disagree, fix report-spec.md before building. Do not ask the authoring agent to choose between conflicting instructions.

If the agent runtime exposes a dedicated session/scratch folder (for example a session-state path injected by the harness), you may also write a copy there for user visibility, but the canonical implementation handoff file remains ./_brief/report-spec.md unless every later authoring step carries the alternate absolute path explicitly.

report-spec.md template

The user-approval doc and agent handoff contract. The Markdown captures sign-off granularity; the embedded YAML captures exact implementation intent.

# Report Spec

## Report identity
- Report name:
- Semantic model:
- Audience:
- Primary purpose:
- Delivery target:

## User decisions and constraints
- Scope:
- Page count:
- Interactivity:
- Design direction:
- Publishing:
- Tooling:
- Model edit permissions:
- Accessibility:
- Data caveats:

## Narrative
- Core story:
- Audience promise:
- Key questions answered:

## Design identity (from `powerbi-report-design` Step 1)
- Tone: <named entry from tone-catalog, e.g. "Editorial Newsroom">
- Signature: <one defining move, e.g. "tabular numerals + display serif headlines">
- Brownfield delta (if applicable): <current_tone → target_tone>

## Page plan (archetypes from `powerbi-report-design` Step 3)
1. Page name
   - Archetype:                      <Executive Summary | Analytical Canvas | …>
   - Layout variant (A/B/C):         <plus one-sentence variant_rationale>
   - Purpose:
   - Visuals:
   - Fields/measures:
   - Slicers/interactions:

## Design system summary
- Theme name + base palette (1–2 lines):
- Color semantics (which measure → which color, 1–2 lines):
- Typography pairing (display + body):
- Layout pattern (grid + gutter + density):
- Accessibility commitments:

## Model requirements
- Existing measures:
- New measures:
- New calculated columns:
- Relationship/sort requirements:

## Canonical design contract

Paste the exact fenced `yaml` block produced by `powerbi-report-design` here.
Do not rewrite it from planner memory and do not replace its mechanical
`layout_contract` with a freeform ASCII wireframe.

The YAML block is authoritative for implementation. `powerbi-report-authoring`
must implement this block; surrounding prose is context and conflict detection.

## Implementation notes

- Model changes:
- PBIR/report authoring:
- Validation:
- Desktop screenshot verification:
- Publishing boundary:
- Risks:

Required acceptance checks before approval

Before writing the approval question, verify the spec meets all of the following. If any check fails, fix report-spec.md and the embedded Design Brief: block before asking for approval.

• The block begins with Design Brief:.
• It includes generated_by: powerbi-report-design and contract_version.
• Every Markdown page has a matching pages[] entry.
• Every page has layout_contract.canvas, layout_contract.grid.regions, and layout_contract.placements.
• Every page has layout_contract.space_audit with empty unplaced_regions and an explicit empty-space/balance rationale.
• Every page has a page_title textbox placement with non-empty title text.
• Slicers are in a top-right filters region or a justified filter rail; no data visual starts under a slicer/header-band region.
• No bare single-value cardVisual is the largest/dominant hero region unless the YAML explicitly describes a composite KPI treatment with context.
• The approved YAML has no ellipses (...), unresolved placeholders, or pages/visuals promised in Markdown but omitted from the YAML.

Approval Gate

After writing report-spec.md, ask exactly one approval question:

Approve this report spec so I can start building?

Recommended choices:

1. Approve — start building
2. Revise audience/purpose
3. Revise scope/page plan
4. Revise design/delivery

Do not invoke authoring or publishing skills until the user approves.

Implementation After Approval

When the user approves, execute this sequence:

1. Re-read the approved canonical report spec (normally _brief/report-spec.md, or the explicitly carried alternate absolute path) and extract the embedded Design Brief: YAML block. Verify it has generated_by: powerbi-report-design, contract_version, one populated layout_contract per page, and space_audit per page before authoring. For greenfield, verify the canvas is FHD (1920 x 1080) unless the user chose another size, and verify the largest/dominant region is not a bare single-value cardVisual.
2. Mark the first implementation todo as in progress.
3. Connect to the semantic model.
4. Create/update required measures and calculated columns using whichever model-authoring path is available — a semantic-model authoring skill, a modeling MCP server, or direct TMDL edits — in that order of preference.
5. Validate each model change with DAX where possible.
6. After calculated-column or measure changes, trigger the lightweight recalculation supported by the chosen tool (e.g., XMLA refresh with refreshType=Calculate) unless a full source refresh is explicitly required and safe.
7. Export model changes to TMDL.
8. If the export writes a flat TMDL layout, reorganize into definition/database.tmdl, definition/model.tmdl, definition/relationships.tmdl, and definition/tables/*.tmdl.
9. Scaffold or copy the PBIP/PBIR report structure.
10. Author PBIR through powerbi-report-authoring guidance.
11. Generate report files.
12. Validate required files and JSON.
13. Open/reload in Power BI Desktop.
14. Screenshot pages.
15. Fix visual, slicer, data-binding, accessibility, and layout issues.
16. Publish only if the approved delivery target includes publishing.

Fabric Publish Rules

Publish only when the approved delivery target includes publishing. Hand the publish step to powerbi-report-management; do not author Fabric REST calls or definition.pbir byConnection payloads from this skill. See powerbi-report-management for the authoritative byConnection schemas (minimal API form and full XMLA form for local Desktop validation), LRO polling, and theme upload rules.

Planner-level rules to respect when invoking powerbi-report-management:

• Resolve workspace, report, and semantic model dynamically; do not hardcode IDs.
• Include all PBIR definition parts on create/update; never send a partial definition.
• Match custom theme upload paths exactly to the paths referenced in report.json.
• Use --verbose for long-running operations and poll LROs to a terminal state (Succeeded / Failed).
• Clean up any temporary publish scripts or payload files after the operation completes.

Validation Standards

A report is not complete until:

• Required PBIP/PBIR files exist.
• All JSON parses.
• definition.pbir points to the expected semantic model.
• Pages and visuals are generated in expected counts.
• Power BI Desktop opens the .pbip.
• Desktop reload succeeds.
• Screenshot capture succeeds for at least the cover and any newly authored pages.
• If published, Fabric LRO returns Succeeded.

Anti-Patterns and Pitfalls

• Generated PBIR is safer than hand-editing individual visual JSON files.
• Model changes must be persisted before Desktop reload.
• DAX filter direction can break dimension aggregations from fact-side flags.
• High-cardinality slicers need search or prefix filtering.
• Desktop validation catches issues that JSON validation cannot.
• Fabric publishing is sensitive to byConnection schema and theme paths.