Improve insight generation prompts with data scientist persona
Details
Updated both cloud and local AI system prompts to frame the AI as a
senior data scientist performing exploratory data analysis. The cloud
prompt now suggests specific section structure (Data Quality, Key
Statistics, Patterns & Trends, Anomalies, Recommendations) and instructs
the AI to use sub-headings, bullet lists, and markdown tables for
well-structured output. Also instructs the model to cite specific column
names, row counts, percentages, and values for concrete analysis.
Desktop 1.55.2 → 1.55.3
Parse markdown headings, bullet lists, and numbered lists in insight pages
Details
BuildSectionBlocks now recognizes:
- ### through ###### headings → heading blocks at the correct level
- Bullet lists (-, *, +) → bullet_list blocks with grouped items
- Numbered lists (1., 2.) → numbered_list blocks with grouped items
Also relaxed chart marker column validation back to analysis columns
instead of chart-eligible columns — the prompt already guides the AI
toward using chart-eligible columns, and strict validation was rejecting
valid view-computed columns that the AI referenced in chart suggestions.
Desktop 1.55.1 → 1.55.2
Fix dismiss button and add View in Notes link for insight cards
Details
Three fixes for AI suggestion cards:
1. Dismiss button now works — AiSuggestionTrayViewModel handles
ContentSuggestionDismissedMessage to remove the card from the UI.
2. After saving insights to Notes, the "Save as Notes" action is replaced
with a "View in Notes" link that navigates directly to the created page.
3. Register INavigationService in DI so the orchestrator can deep-link
navigate to the insight page when the user clicks "View in Notes".
Desktop 1.55.0 → 1.55.1
Include dataset_id value in aggregate error messages
Details
When Uuid::parse_str fails on a dataset_id, the error message now
includes the actual value that failed to parse, making it much easier
to diagnose why chart data loading fails. Previously only returned
a generic "invalid dataset id" with no context.
Fix chart error swallowing in aggregate queries
Details
DatasetService.AggregateAsync and AggregateGroupedAsync were silently
returning empty results when the Rust FFI returned errors, causing charts
in AI-generated insight pages to show "No data available" instead of the
actual error message. Changed both methods to throw
InvalidOperationException with the error details so ChartDataLoader can
surface them in the UI. Added ExtractErrorMessage helper to parse the
error from Rust JSON responses. Also added debug logging for aggregate
queries, normalized aggregation values to lowercase in
TryParseChartMarker, and strip quotes from chart title values.
Version: 1.54.4 → 1.54.5
Get notified about new releases