Release Notes

Highlights

• Grouped tables and relationship summaries - group Records by up to four properties or relationships, expand groups to their Records, and add count, sum, average, minimum or maximum columns over related Records.
• Queries shaped for your application - return selected fields, aggregates, nested objects and samples of related Records from the API, SDK, CLI or an Object - Query node.
• Editable query rows - bind Inputs and a Save button to identifiable Records in a list without adding an entity context for every row.
• Repeating events and periods - store recurrence rules with time zones, skipped dates and optional durations, then find occurrences in a Flow.
• Document previews and number formats - show documents in Experiences and choose how numeric values appear without changing the stored values.
• CLI file transfers - list, download and upload attachments, or create a Record with its required files in one command.

Schemas and Records

• Grouped views. Group by text, number, boolean, date or duration properties, or by a relationship. Date grouping supports day, week, month, quarter and year. Each group shows its Record count and the totals of visible number and duration columns. Expand groups to work with their Records, use Shift-click across expanded rows, and save the grouping with a view.
• Relationship groups. Group by the linked Record, shown by its name. A Record linked to several Records appears in each group. Order groups by a property of the linked Record, such as a Stage's order. Records with no readable link appear under (none).
• Aggregate columns. Add count, sum, average, minimum or maximum columns over a relationship, with an optional filter on the related Records. Sort the Records table by an aggregate column across pages. Counts and totals include only Records the viewer can read; No access is distinct from a zero result.
• Recurrence fields. Store repeating events or periods, such as weekdays at 09:00 or the last Friday of each month. Configure an interval, time zone, optional start and end, occurrence count, skipped dates and duration. Local times are preserved across daylight saving changes.

Recurrence fields available in the Table, Record, and Experiences view for complex repeating events.

• Number formatting. Choose integer, decimal, percentage, compact or scientific presets, or a custom decimal, percentage or currency format with decimal-place and separator settings. Formats affect display, not stored values. Currency selection starts with AUD.
• Relationship lookup. Record selectors support keyboard navigation and a mobile-friendly tray, making the selected value easier to inspect and keeping dropdowns within the page.

Table view showing a Grouped view, aggregate columns, number formatting (AU Currency).

Experiences

• Editable rows. Inputs inside a For Each over query results can edit the row's Record, and an explicitly targeted CRUD Save or Update button can save it. A selected-column query must include id. Aggregates, group keys and standard properties remain read-only. Unsaved edits to the same Record are shared across the Experience, and saving sends only changed fields. Each row needs its own save action; saving several rows from one button is not supported.
• Shaped queries. Query blocks can read selected fields, relationship aggregates and nested query output. Bar charts and word clouds can query grouped Records directly, including configurable date labels for bucketed keys.

Experience showing bar charts querying against grouped records.

• Document previews. The Visualisation block's Document option previews PDFs, Word documents, PowerPoint presentations and images from a URL or Record file. It supports an optional download button. Unsupported files or failed previews appear as links; a multiple-file source previews its first file.

Experience shows a PDF document preview that is loaded, visualised, and downloadable.

• Number formats in content. Numeric variables can override the field's default number format when inserted into content.

Experience shows different number formats (AU Currency, Percentage, etc).

• Draft preview. Studio offers Preview for a draft Experience version and View for a published one. Preview uses the selected version's path, even when it differs from the published path.

View for the Published Experience version.

Preview for the unpublished Draft Experience version.

• More reliable first renders. Entity contexts looked up by Record ID include the Record in the initial page render. Missing or unreadable data no longer prevents the rest of the page from rendering. Date defaults use one consistent time for the render, avoiding changes between the initial page and its interactive state.

Flows

• Recurrence node. Data - Recurrence lists occurrences in a date range, finds the next or previous occurrence, or checks whether a date-time falls on an occurrence or within its duration. The same operations are available as expression functions.
• Flow direction. Switch a Flow between top-to-bottom and left-to-right, rotate selected nodes independently, and set a preferred direction for new Flows. Layout changes support undo and redo.

Example of a left-to-right horizontal Flow direction.

• Whole-graph authoring. Create or update a Flow graph through the API or CLI without supplying generated IDs or node positions. Refer to connected nodes by their nodeId; optional auto-layout arranges the graph.
• Broken-connection recovery. Saving a Flow lists invalid connections with their nodes and handles. Choose whether to remove them and save, rather than repairing the saved data manually.

API, SDK and CLI

• Record queries.records.query in the SDK and ligantic records query expose selected fields, aggregates, grouping and nested output. Include id explicitly when the returned rows need a Record identity.
• File attachments.ligantic records files list, download and upload manage Record attachments. Downloads stream to a file or stdout. records create --file <field>=<path> creates a Record with files; its create Flows run after the uploads are confirmed.
• JSON from files or stdin. Every JSON option accepts inline JSON, @<path> or - for stdin, including filters, patches, Flow inputs and generated command options. Only one option per invocation can read stdin; invalid JSON fails before a request is sent.
• Searchable pickers. Interactive selection lists filter as you type and fit the terminal's height.
• Record versions. Record reads and updates return the Record's version so clients can track which version they have received.
• Rate-limit backoff. Rate-limit responses include Retry-After, and the SDK and CLI honour it when retrying eligible requests.

Fixes

• Grouping a Records table keeps its existing filter and sort; expanded groups stay open across navigation, and totals respect the property's number format.
• View selectors scroll instead of overflowing. Dropdowns stay on-screen on tall pages, and choosing a Record in a Flow input popover no longer closes the popover unexpectedly.
• Inline creation links a related Record once, respecting the relationship's cardinality. A name-based write resolves correctly when a field's ID is also another field's name.
• Record counts remain correct when requesting a count without rows. An empty or filter matches no Records.
• A Flow run's initiator can download its output files. PDF capture waits for Experience data to settle before taking the snapshot.
• Experience links correctly substitute parameter placeholders. An inaccessible Space returns a not-found page rather than an internal error.
• Google OAuth profile callbacks are no longer incorrectly blocked.

Changes to Be Aware Of

• Relationship fields return only links to Records the caller can see. Saving a relationship selection preserves links the caller cannot see; explicitly linking to an unreadable target is rejected. Required relationship validation uses the caller's visible selection.
• records put now synchronises relationship fields supplied in its data. Relationship fields omitted from the request keep their existing links, although omitted ordinary data fields are still reset by replacement.
• A relationship nested inside an array must be changed on a specific array row through a Record write. The standalone link endpoint cannot address it and rejects the request.
• Sorting by an aggregate column is not applied while searching a Records table. Grouped views have a maximum of four keys and 100 groups per page; search is not applied to groups.
• File fields imported from JSON Schema without a size limit now default to 5 MB.