teable

Feature Updates

• Improved Teable Agent resource link recognition: Agent now better recognizes Teable table, app, automation, and record links.

Bug Fixes & Optimizations

• Fixed organization permission inheritance: Members now correctly receive permissions from departments or user groups, including parent departments.

• Fixed date formula evaluation issues: IS_BEFORE, IS_AFTER, and IS_SAME now return correct booleans in logical functions.

• Improved status/date formula stability: Time-based process states are now more reliable.

• Fixed record insertion after attachment upload: Reduced “Attachment not found” errors when inserting records after uploads.

• Fixed inaccessible or undeletable error tables: Tables in error states can now be removed for easier recovery.

• Improved chat response speed: Cached information is prioritized to reduce wait time.

• Improved chat attachment handling: Files sent through guided input are now preserved, displayed, and passed into chat.

• Fixed attachment upload blocking in queued chats: Improved upload stability in queued chat scenarios.

• Fixed duplicate AI replies: Reduced duplicate responses after the first instruction.

• Fixed aggregation result loading failures: Tables with mixed-case field names now load summary results more reliably.

• Fixed saving shared Bases as copies: Users can now save copies when shared links allow “Save as copy”.

• Improved app card display in AI conversations: Related app cards now appear below messages with context preserved.

• Improved CLI help and documentation: Clarified scope, base parameters, skill import, GitHub skill paths, login config, and SQL table naming.

• Improved Agent Computer startup experience: Enhanced app preview prewarming for faster, more stable app opening.

• Unified SQL query experience: SQL query permissions are now consistent across UI and CLI.

• Improved table content display in AI Chat: “Add to chat” now shows field names instead of column positions.

• Improved AI Chat selection markers: Row and column selections are clearer, helping AI understand referenced ranges.