Bind fields
Build a privacy-aware Field Manifest from the current visual query.
Combine AI-assisted JSON generation, reusable logical datasets, controlled interactions, declarative SVG, Leaflet and ArcGIS maps, and stable targeted changes—without turning dashboard content into executable code.
HyperPBI composes local prompts and never calls an AI API. The user copies reviewed context to an approved service, pastes one response, and explicitly controls preview and apply.
Build a privacy-aware Field Manifest from the current visual query.
Author directly or copy a governed AI prompt.
Prepare the complete candidate without replacing valid state.
Locate rendered content by canonical authoring path.
Review the mutation summary, then apply explicitly.
Preserve stable IDs, Power BI selection, and filter safeguards.
Inspect preview mode maps a clicked runtime component to its editable authoring owner. Candidate edits are prepared and validated before the preview changes; an invalid edit keeps the last valid dashboard intact.
regional_performanceregionalSummary/components/0/children/2Authoring nodeThe real Inspector uses a resizable two-pane desktop layout and an accessible Tree / Properties switch at narrow Power BI visual sizes. Stable IDs support targeted AI replacement while unrelated nodes remain unchanged.
Build KPI illustrations, process flows, pumps and equipment, and infrastructure diagrams from allowlisted SVG elements.
svgMarkup escape hatch when structured SVG is insufficientforeignObject, images, or external SVG resourcesPut every Power BI field in the single Values well, then configure each layer's dataset and source bindings in canonical JSON. Logical datasets create different layer views over the one flattened data view received from Power BI.
none, explicitly configured, user-triggered, and policy-gatedHyperPBI is an analytical Web GIS builder for Power BI—not a feature-editing or 3D GIS replacement.
The inventory, maturity, Inspector controls, strict properties, field handlers, documentation, AI reference, and valid examples are generated from canonical component descriptors.
Every advanced surface has a parser, allowlist, scope, and limit. Dashboard JSON never gains unrestricted code execution or provider access.
No eval, functions, callbacks, inline handlers, SQL, or network datasets.
DOMPurify plus parsed, allowlisted, selector-scoped CSS and URL policy.
Allowlisted elements and references, namespaced IDs, and exact rendering budgets.
Fields, samples, masked, summary, or types modes remain local until copied.
External filters require a real Power BI model-column target.
Independent tile, geocoder, and ArcGIS service authorization with normalized errors.
Both packages expose the same Power BI data-role contract. Their difference is WebAccess and provider availability—not authored schema behavior.
All non-network dashboard features and bound Power BI map data.
npm run package:coreConfigured HTTPS tiles, geocoders, and ArcGIS services subject to host policy.
npm run package:mapsHyperPBI schema version 2.0 is an authoring contract. PBIVIZ package version 1.0.0.0 is the Power BI artifact version; they evolve independently.
Generated references are checked against canonical descriptors and checked-in source guidance.