Version 0.5.0
stable
A major Sample Data Studio upgrade with realistic generators, inline templates, guided lessons, task-local session variables, and schema-shaped AI assistance, alongside expanded SL, Views, Topic Flow, AI provider, startup, settings, and packaging improvements.
What changed
- Added clearer SL validation messages.
- Added AI-assisted SL repairs.
- Expanded string filtering with containsAll(), multi-value startsWith() and endsWith(), blank checks, literal replace(), and safe like() wildcards using * and ? instead of regular expressions.
- Added array filtering helpers with countWhere() and none(), plus sum(), avg(), min(), and max() numeric aggregates.
- Added coalesce() for schema evolution and get()/hasKey() for exact JSON or header keys containing dots, slashes, or other punctuation.
- Updated SL autocomplete and the reference guide with practical anonymized data samples and examples.
- Updated both Claude and ChatGPT/OpenAI filter generation and repair with the same expanded SL reference and sample-guided instructions.
- Delayed the initial server connection failure message while displaying a connection spinner during startup.
- Added a View tab to message details for displaying selected fields vertically as name/value pairs.
- Added reusable SL computed fields to the result-view customization manager; values are evaluated for the selected message and appended to its View without adding duplicate message-detail tabs.
- Added generic JSON name/value field mappings to custom result views, supporting arbitrary flat or nested selectors and literal field names containing dots.
- Added local JSON hierarchy suggestions and a Determine from JSON action, while retaining automatic key/name-to-value discovery and typed-value unwrapping without sending message data to AI.
- Unified the selected message-details tab between side- and bottom-docked layouts.
- Expanded AI provider settings with No AI mode plus Gemini and OpenRouter, including provider-specific keys, model discovery, and the same criteria, topic-filter, product-assistant, and playbook-assistant capabilities as ChatGPT and Claude.
- Added searchable AI model selectors and compact, copyable provider errors so long API responses no longer expand the Settings layout.
- Rendered AI assistant replies as safe GitHub-flavored Markdown with formatted headings, lists, tables, links, and code, plus a one-click copy action.
- Added persistent Topic Flow diagrams for Views with draggable Kafka topics and process nodes, process types and environments, descriptions, colors, resizing, directional links, locking, minimap navigation, and cross-window-safe editing.
- Added live Topic Flow telemetry with filter and cleanup-policy indicators, received and shown counts, animated traffic intensity, and View live controls available across the Messages and Topic Flow tabs.
- Constrained Settings to 80% of the application viewport with a styled scrolling content area and added configurable Topic List size and record-count refresh intervals.
- Added clearer restart-to-update guidance on Windows while the application update is being installed.
- Improved Linux release packaging with the installation helper included in uploaded artifacts.
- Added AI-assisted computed-field generation for custom message Views, with selected-message context, SL validation, generated labels, and live previews.
- Added an AI Assist tab to Data Samples that can generate and apply coordinated sample names, session variables, headers, keys, and messages using the complete replacement-function catalog and starter templates.
- Expanded Data Samples with a correlated Things (Large) generator covering vehicles, household appliances, clothing, countries, codes, industries, science, all 118 chemical elements, and broad financial data.
- Added coherent financial sample generators for 50 currencies, FX pairs, exchanges, equities, bonds, commodities, crypto assets, transactions, balanced journal entries, and reconciled financial statements.
- Added primitive-only Inline variants beside every Person, Order, Invoice, Payment, Shipment, Product, Device, Organization, Account, Subscription, and Support Ticket template.
- Added an alphabetized Lesson template group demonstrating random choices, identifiers, timestamps, counters, references, stateful session variables, nested conditions, and a complex network-testing event.
- Improved Data Samples AI Assist to construct realistic templates from pasted JSON, schemas, or field shapes, explain dependent fields with @@if examples, and prioritize Inline and Lesson patterns for AI and Skills context.
- Expanded replacement autocomplete and help for the new Things and financial namespaces, with examples that cross-reference session variables, headers, JSON keys, and message fields.
- Added isolated backend Sample Loader sessions with parent and child variable merging, stable and recalculated values, nested references, autoInc, bounded wrapping @@inc, autoCycle, autoDateInc, and successful-write-only state advancement.
- Added a conditional Sample Loader notice explaining that server-side session state advances only for the task and is not written back to the saved sample or editor.
- Made Sample Loader target topics intentionally blank by default with an explicit protection watermark, preventing accidental reuse of the current or previous topic.
- Improved message-details docking controls and kept Playbook AI assistant actions compact when displaying long tool lists.
- Task errors were not being shown in the tasks modal
- Make sure @@count service side function is represented in data sample screen when test button is pressed with 1st value