What's New
122 updates shipped in 52 days • 2.3 updates per day
We emphasize speed of updates (while being safe and secure) to make Agents for Data as good as possible for you.
Note: Agents for Data might not have all features yet, but we're moving very fast to add them.
Coming Soon
We released Logs (Beta) so that you can view and analyze application logs in real-time with automatic console log capture, filtering by severity and search capabilities, helping you debug issues faster with structured log data
We released Performance Monitoring (Beta) so that you can track and improve your app performance with automatic distributed tracing, helping you identify bottlenecks and optimize operation times
We redesigned the datasets list so that you can browse many datasets more easily with a searchable table view, fuzzy search that highlights matches, and smart sorting that shows recently viewed datasets first
We improved query execution speed so that queries run faster with reduced wait times, especially during high traffic periods when many queries are running simultaneously
We optimized data ingestion so that batch inserts are significantly faster and no longer slow down other queries during high ingestion rates
We fixed memory issues in query execution so that the system stays stable and responsive even when processing many queries over time
We added automatic console log capture so that browser console logs are automatically forwarded to your logs viewer for easier debugging without requiring manual log statements
We released Product Analytics (Beta) so that you can track user events and properties, create analytics environments with automatic dataset provisioning, and analyze user behavior with flexible event tracking
We reorganized apps navigation so that Agent Forms and Product Analytics are grouped together in an expanded Apps panel, making it easier to discover and access different apps
We refocused Product Analytics messaging so that it emphasizes product team value with clear benefits around feature adoption, conversion funnels, and data-driven product decisions
We improved Product Analytics data integrity so that user properties are stored cleanly without anonymous user IDs, ensuring better data organization and more accurate user tracking
We fixed upload button being incorrectly disabled so that you can upload files immediately without waiting for usage data to load
We fixed "Data Viewer" title breaking across multiple lines so that it displays cleanly on all screen sizes
We released Artifacts so that you can create, save, and share charts (datagrid, line, bar, scatter) as shareable artifacts with public or private visibility, fuzzy search to find them quickly, and a mobile-friendly editing experience. Dashboards, reports, and SQL queries coming soon
We added fuzzy search to artifacts so that you can find charts and artifacts even with typos or partial matches, with highlighted search results
We improved mobile artifact editing so that controls are more accessible with better scrolling and an optional fullscreen drawer for easier editing on phones and tablets
We improved chart reliability so that charts with table names containing special characters like hyphens load and edit correctly without errors
We improved chart loading so that charts use dataset information from the backend instead of parsing SQL, making them more reliable and faster to load
We updated connector release dates so that Stripe, Google Search Console, and Google Ads connectors appear in the "Coming Soon" section with November 2025 release dates
We fixed flickering "no artifacts" message so that the artifacts page loads smoothly without brief error messages appearing during refresh
We reorganized navigation so that less-used tools (SQL, View, Convert, What's New, Feedback) are now in a "More" overlay panel, keeping the main sidebar cleaner and easier to navigate
We improved hover panel stability so that overlay panels stay open longer and are easier to interact with when moving your cursor
We added dataset settings so that you can configure public data ingestion for forms, webhooks, and IoT devices without requiring authentication
We improved dataset protection so that system-managed datasets (like Stripe sync data) are automatically protected from accidental modification
We reorganized dataset page buttons so that Settings and Share are separate, making it easier to find configuration options
We fixed mobile layout on the Agent Forms marketing page so that it displays correctly on phones and tablets without horizontal scrolling
We improved marketing page hierarchy so that the main value proposition is more prominent and easier to understand
We improved marketing copy to be more engaging and less technical, making Agent Forms more accessible to non-technical users
We fixed SQL query errors so that tables with special characters like spaces and parentheses work correctly without causing 500 errors
We fixed table navigation so that dataset tables with special characters in their names load correctly instead of showing "Table not found" errors
We improved form links so that they display the full URL and include a direct "View Results" button to jump straight to your data table
We improved social media previews so that shared Agent Forms links display proper preview images with marketing content
We added AI chatbot assistant for forms so that you can fill out forms conversationally with an AI guide, and answers automatically populate the form fields
We improved form list UI so that form links are visible directly on each card with copy buttons and quick-open actions, making it easier to share and access your forms
We improved form table naming so that tables are named after form titles instead of random IDs, making it easier to identify which table belongs to which form
We replaced external feedback links with an integrated feedback modal so that you can share feedback directly without leaving the app
We improved connections page so that marketing content shows when logged out, helping visitors understand what connections are available
We improved agent forms demo so that transitions between chat and form steps are smoother and show both views side-by-side
We released Agent Forms so that you can create intelligent web forms that collect structured data and store it directly in your workspace datasets
We improved query reliability when using quoted workspace and dataset names so that all SQL queries execute correctly regardless of naming format
We enhanced SQL processing to handle various naming conventions so that your queries work whether you use quotes or not
We fixed data table caching so that switching between datasets always shows the correct table instead of displaying the previous dataset you viewed
We fixed pagination in wide datasets so that loading more rows works correctly even when scrolled horizontally, eliminating the issue where tables with many columns wouldn't load additional rows
We fixed dataset loading so that workspace information loads correctly and dataset pages work reliably
We fixed datagrid pagination so that scrolling through large datasets no longer jumps back to the top or resets to the first 100 rows
We improved scroll position stability so that loading more rows happens smoothly without interrupting your browsing
We added infinite scroll to table viewer so that you can browse through millions of rows by simply scrolling, without clicking pagination buttons
We improved table loading speed so that initial page loads instantly with just the first 100 rows instead of loading thousands at once
We added automatic data loading so that more rows appear smoothly as you scroll down without any extra clicks
We fixed data display during sorting so that you see consistent results without mixed old and new data
We made loading overlays more subtle so that they provide feedback without being distracting
We optimized tables with 100+ columns so that they load and scroll dramatically faster by only rendering columns you can actually see
We fixed column width preservation so that your manually resized columns stay the same width when you sort or filter data
We added fuzzy search to schema results so that you can quickly find columns in tables with many fields by typing partial matches
We improved the XY plot data modal so that it uses 95% of your screen width, letting you see multiple columns at once instead of being limited to a narrow view
We added shift+scroll support to data tables so that you can easily scroll horizontally through wide datasets
We added collapsible columns to schema results so that you can hide column details and focus on the columns you care about
We improved payment forms so that the submit button stays disabled until all payment details are complete, preventing frustrating "card number incomplete" errors
We improved usage page loading so that content appears with smooth skeleton layouts instead of spinning indicators
We implemented grandfathered file limits so that existing users' files created before October 14 don't count toward limits for 2 months, giving you time to adjust
We improved table schema display in chat so that column types appear in plain language (text, number, date) instead of technical database terminology
Improved table loading experience with smooth overlays instead of skeleton screens - no more visual jumps when data loads
Added server-side sorting for dataset tables - much faster performance with large datasets
Fixed loading indicators to show clearly when sorting or filtering tables
Updated homepage title for better search engine visibility
Released MCP server (Beta) so you can connect Cursor and Claude Desktop directly to your workspace - ask AI to browse your datasets and run SQL queries without leaving your code editor
Added usage limits visibility - see exactly what your plan includes and current usage
Fixed file count display to accurately show both uploaded and cloud storage files
Enhanced file conversion for free tier users - sequential processing prevents errors
Improved Microsoft account integration for Google Workspace connections
Better usage tracking for more accurate credit and limit calculations
Updated interface language to be clearer and more user-friendly
Launched new 4-tier pricing with clearer limits and 3-hour AI credit refresh cycles
Added file limit enforcement - Free tier includes 5 files, upgrade for unlimited
Added Claude AI model restrictions for Free tier (Claude Haiku and Sonnet 3.5 available on Pro+)
Real-time usage tracking shows current usage against your plan limits
Added prominent file limit warnings with easy upgrade options throughout the app
Fixed workspace loading issue that prevented new users from accessing their workspace after sign-up
Improved compatibility with AI development tools for seamless integration
Generous free tier upgrades - increased file conversions from 10 to 20 per day, storage expanded from 50MB to 500MB, and removed file count limit (now unlimited)
Improved AI credit system infrastructure for more accurate usage tracking and better performance monitoring
Major rebrand: TabLab is now Agents for Data! New name, new domain (agentsfordata.com), and refreshed brand identity that better reflects our focus on AI-powered data analysis
Complete design system refresh for comfortable all-day use - softer colors that are easier on your eyes, improved contrast for better readability, and a more polished professional look throughout
Fixed data editing - cell edits now save immediately and persist when scrolling or filtering
Improved chat experience with smart scrolling that doesn't interrupt reading and always-accessible input field
Enhanced data analysis agent with automatic SQL error detection and self-correction for more reliable queries
Added editable chart titles with auto-save - customize chart names directly in conversations
Enhanced chart creation with interactive visualizations (line, bar, scatter, data grid) directly in chat
Cleaner homepage design with improved pricing layout and clearer messaging about built-in AI agents
Better dataset preview styling with reduced visual noise and more professional appearance
Fixed dataset uploads to work more reliably - tables now consistently visible after creation
Added PWA and Windows app users now land directly on home dashboard instead of marketing page
Improved SQL page user experience - table action buttons moved to left for better visibility with long names
Enhanced dataset list to automatically refresh after creating new datasets (no manual refresh needed)
Improved data analysis agent reliability with simplified table discovery and more consistent SQL generation
Added comprehensive usage panel in sidebar showing AI credits, file conversions, storage, and team usage (A/B test)
Fixed sidebar layout issues and improved navigation organization with proper section headers
Fixed dataset preview functionality - previews now work correctly even without table metadata
Improved dataset viewer error handling with retry functionality and better error recovery
Enhanced pricing UI feature flag controls - pricing banners only appear when feature is enabled
Fixed navigation and React hooks violations for better app stability and performance
Enabled Data Analysis agent for interactive data exploration and insights
Activated high-performance data lake operations for faster analytics
Enhanced subscription status checks with more reliable error handling
Introduced comprehensive usage limits system with clear warnings and transparent tracking
Improved billing accuracy by tracking actual file sizes instead of estimates
Enhanced user experience with informative error messages when usage limits are reached
Added granular usage tracking for better subscription value and resource management
Faster data processing with server-side query execution for large datasets
Added professional preview images for all marketing pages when sharing on social media
Improved social sharing with proper Open Graph images for Twitter, LinkedIn, and Facebook
Fixed dataset viewer to work correctly with all dataset types
Faster data loading - queries now run server-side without downloading large files
Added ability to query multiple tables and datasets simultaneously
Fixed automated background tasks (imports, syncs, cleanups) to run reliably
Improved 404 error pages with better design and helpful navigation
Added "What's New" feature in sidebar to stay updated with latest improvements
Enhanced Open Graph images with actual page screenshots instead of generic text
Improved social media previews to show attractive hero sections
Removed AI credit system - launching new pricing limits A/B test
Fixed table sorting functionality in dataset viewer - columns now sort correctly
Added automatic preview images for links shared on social media and messaging apps