Changelog

Changed

• Grid: Boolean column filters now show True, False, Null, Not Null options instead of generic text filter. #1028
• Transactions: DDL statements (ALTER, CREATE, DROP) are now tracked as uncommitted changes when auto-commit is OFF on databases with transactional DDL (PostgreSQL, SQL Server, Snowflake).
• Transactions: Non-transactional DDL drivers (MySQL, Oracle, DB2) now warn before executing DDL with auto-commit OFF, even with no prior uncommitted changes.
• Roles: Added autoCommit setting to connection roles, allowing per-role auto-commit defaults that override the global defaultAutoCommit setting.
• Tab Titles: Added ${role} variable to dbcode.tabTitle.format for showing the connection role in tab titles.
• Results: Binary cell “Save as File…” context menu to download cell contents to disk with auto-detected file extension.
• Results: Binary cell “Open in Editor” to view and edit binary data in VS Code editors (hex editor, text, etc.) with round-trip saving back to the grid.
• Bunny Database: Added support for Bunny Database (bunny.net) as a cloud libSQL service. #1040
• MongoDB: Added support for multiple hosts in replica set connections without SRV. #1005
• MySQL/MariaDB: Added “Allow Local Infile” option to enable LOAD DATA LOCAL INFILE statements.

Fixed

• Grid: Fixed image preview in binary cells blocking the edit input when clicking to edit.
• Tree: Clicking a table in the tree now keeps focus in the tree, matching VS Code’s native behavior. Enter opens and focuses the table. #1039
• MCP: Fixed dbcode-disconnect tool not working in Copilot Chat (missing tool registration). #1034

Changed

• Monitoring: Preview of server monitoring, with support for Postgres, SQL Server, Oracle, and others.
• AI Assist: Custom model errors now show the provider’s error message instead of a generic status code. Token usage (prompt, completion, total) is displayed after each custom model response.
• MCP: Added dbcode-disconnect tool to allow AI agents to programmatically disconnect database connections and release file locks. #1034
• Results: Added result tab rename via double-click or right-click context menu. Custom names persist across re-executions when the SQL matches. #1033
• Tables: Single-clicking a table reuses the previous tab, double-clicking pins it. Tabs are protected from reuse when editing data, filtered, sorted, streaming, or loading more rows. #1036
• Tree: Ctrl+C / Cmd+C copies selected item names (supports multi-select). F4 opens Script DDL for the selected item. #1039
• Tree: Column tooltips now show type, nullable, primary key, default, FK reference, and comments. #1039

Fixed

• Editor: Fixed output log and tab badges not showing affected row counts for DML statements. #967
• SQL Server: Added FOR SYSTEM_TIME temporal table syntax support in language server. #1027
• Grid: Fixed paste inserting clipboard contents twice on Linux. #1024
• Grid: Fixed two context menus appearing when right-clicking in relation detail grids. #1025
• Export: Fixed Excel export shifting dates back by one day for users in timezones west of UTC.
• DuckLake: Fixed Google Cloud Storage connections getting 403 errors. #1031

Changed

• PostgreSQL: Added Microsoft Entra ID authentication support (SSO, CLI, and Service Principal).
• MySQL: Added Microsoft Entra ID authentication support (SSO, CLI, and Service Principal).

Fixed

• Editor: Fixed autocomplete not providing suggestions when SQL comments precede a statement in Postgres.
• CSV: Fixed file/folder mode not persisting when reopening a CSV folder connection.
• MySQL: Fixed information_schema and other system schema views not appearing in the object browser.

Changed

• SAP ASE (Sybase): Added in preview, requires Java.
• Google Cloud Spanner: Added in preview with support for ADC, Service Account, and Emulator authentication.
• Notebook Export: Python cell outputs (DataFrames, plots, print output) are now included in exports.
• Notebook Export: Python code blocks now have syntax highlighting.

Fixed

• Notebook Export: Fixed PDF and HTML exports not downloading.
• Snowflake: Unified SSO and password auth paths so MFA/SSO only prompts once per connection. #1019
• Grid: Fixed relation icon resolving the wrong foreign key when a different cell was focused. #1022
• Snowflake: Fixed Go to Definition failing for tables and views in non-default schemas by using schema-qualified names in GET_DDL calls. #1021

Changed

• Export: Added progress overlay with status messages and cancel button during file exports. #980
• PostgreSQL: Added progressive schema loading for faster connections to databases with large schemas. #894 #878
• Explorer: Databases within a connection are now sorted alphabetically, with the configured database pinned to the top. #923

Fixed

• Redshift: Fixed database dropdown only showing the connected database instead of all databases.
• BigQuery: Script CREATE (DDL) for stored procedures now includes IN prefix for input parameters, matching BigQuery Console output. #1000
• SQLite/LibSQL: Fixed computed (generated) columns not appearing in the explorer.
• IBM i: Fixed missing tables, views and tunnel routing issues. #976

Fixed

• Snowflake: Fixed “terminated connection” errors during introspection by validating pooled connections before use. #998
• MySQL: Fixed full compatibility with ShardingSphere-proxy. #987
• Trino/Athena: Progressive schema loading is now enabled by default, preventing a single table with corrupted metadata from blocking introspection. #997
• Favorites: Fixed favorites panel hanging indefinitely on startup. #995
• DB2: Fixed DML statements (INSERT, UPDATE, DELETE) always reporting “0 rows affected”. #996
• Elasticsearch: Fixed connection failure on Elasticsearch versions prior to 7.16 due to unsupported SHOW CATALOGS command. #994

Changed

• IBM I: Added preview support for IBM I (AS/400) databases. #976
• SQLite: Added sqlite-regex extension (PCRE2 regular expressions). #992
• SQLite: Enabled load_extension() for loading custom extensions via post-connection SQL. #992
• Excel Export: Replaced SheetJS with custom XLSX writer using Shared String Table deduplication, producing ~60% smaller files. #980
• Export: All grid exports now use the native VS Code save dialog with Open File/Open Folder/Copy Path actions. #796

Fixed

• MySQL: Fixed compatibility with ShardingSphere-proxy and other MySQL proxies that don’t support IFNULL. #987

Changed

• Query Builder: Visual query builder for constructing SELECT queries with a drag-and-drop. Add tables, configure joins, select columns, set filters, GROUP BY, HAVING, ORDER BY, DISTINCT, and LIMIT. Includes AI assistance for natural language query building, Save/Save As to Favorites.
• AI: AI Assist panel for results. Manipulate query results using natural language. Filter, sort, group, pivot, chart, and more without touching the grid UI.
• Favorites: Favorites folder now appears at the top of database nodes in the connection tree, showing favorites tied to that database. #982

Fixed

• Results: Fixed grid jumping to the first column when clicking a column header while scrolled right. #988
• Export: Excel export performance improved ~2.5x and file size reduced ~60% using compression.

Fixed

• Results: Fixed clicking in the grid sidebar (Inspector, Columns, etc.) stealing focus from the panel.
• Results: Fixed Saved Filters popup not appearing when clicked.

Changed

• Editor: Changed execute keyboard shortcut from Cmd/Win+Enter to Ctrl+Enter for cross-platform consistency.
• Streaming: Added Firebase Firestore snapshot subscriptions for real-time collection changes. Right-click a collection → Subscribe.
• Streaming: Added RavenDB Changes API subscriptions for real-time document changes. Right-click a collection → Subscribe.

Fixed

• Editor: SQL Statements are executed individually when multiple statements are selected, allowing faster result returns, error handling, and correct tab identification. #967
• PostgreSQL: Error messages now include additional context from the database (where, detail, hint) when available.
• SQLite: Fixed attached databases showing as “Default” after refreshing when a non-main database was last browsed. #984
• RavenDB: Fixed error before connection is established. #986
• Results: Clicking empty space around rows now focuses the grid for keyboard navigation. #985
• AI: Copilot model selection no longer auto-selects the first available model which could silently consume credits on expensive models. Now uses Copilot’s ‘gpt-4o-mini’ family by default as recommended by CoPilot for completion. Removed the dbcode.ai.modelId setting.
• Connections: Fixed saving connections erroring in some cases.

Changed

• Streaming: Real-time message streaming for Kafka topics and RabbitMQ queues with LIVE indicator, event counter, throughput stats, and pause/stop controls in the grid status bar.
• Streaming: Added PostgreSQL LISTEN/NOTIFY channel subscriptions via LISTEN channel_name; or right-click database → Subscribe.
• Streaming: Added MongoDB Change Streams for collection subscriptions via db.collection.watch() or right-click collection → Subscribe. Requires a replica set or sharded cluster.
• Streaming: Added Redis Pub/Sub channel subscriptions via SUBSCRIBE channel_name or right-click database → Subscribe.
• Streaming: Added SurrealDB LIVE SELECT table subscriptions via LIVE SELECT * FROM table; or right-click table → Subscribe.
• Active Streams: New view, showing all running streams grouped by connection.
• Editing: JOIN query results are now editable. Columns unique to one table in the query are editable, while columns present in multiple tables are read-only. Primary keys are resolved automatically, or you can select key columns manually when saving.
• Transpose: Field column is now sortable, click the header to sort field names alphabetically. #983
• Results: Added “Clear This Column Sort” and “Clear All Column Sorts” options to the column header menu.

Fixed

• Connections: Fixed stale connection error after dropping a database and refreshing. Refresh now cleans up connections for databases that no longer exist and avoids redundant database list queries. #966
• MCP: Fixed autoStart not reliably starting the server on VS Code reload. #960
• Export: Replaced Results Excel export, fixing non-numeric columns (VARCHAR, UUID, etc.) being formatted as numbers, improving file size and performance for large exports, and adding the SQL query as a second “Query” sheet. #980
• Import: Fixed driver name mapping during JSON import causing names like “sqlserver” to not map to “mssql”. #979
• Results: Fixed Cmd/Ctrl+C not copying grid selection when a WHERE filter is active.
• Results: Fixed horizontal overflow causing page-level scrolling.
• ERD: Fixed saved node positions not being restored when reopening a diagram. #829
• Redis: Fixed connection failing on managed Redis services (Google Cloud Memorystore, some AWS ElastiCache configurations) that block the CLIENT SETNAME command.
• Language Server: Fixed ClickHouse parser not recognizing CTEs in INSERT…SELECT statements. #981
• Transpose: Improved performance when entering transpose mode with large result sets. #983

Changed

• Apache Kafka: In preview with topic browsing, JSON auto-expansion, message producing, consumer group visibility.
• RabbitMQ: In preview with queue browsing, message peeking, exchange topology with bound queues, message publishing.
• Valkey: Added Valkey as a Redis derivative with its own icon and connection type. #970
• AI: Inline completion now scopes context to the active statement instead of the entire document, reducing noise and improving suggestion quality.
• Results: Changed duplicate row shortcut from Ctrl/Cmd+D to Ctrl/Cmd+Shift+D to free Ctrl/Cmd+D for fill-down (matching Excel/Sheets convention). #972
• Editor: Active statement highlighting now includes leading and trailing comments.

Fixed

• Azure SQL: Fixed stale Entra ID tokens requiring manual sign-out by automatically re-authenticating when token is rejected.
• Azure: Fixed cloud provider connections failing with ENAMETOOLONG when Azure resource IDs are deeply nested. #963
• MySQL/MariaDB: Fixed long-running queries failing with ER_SOCKET_UNEXPECTED_CLOSE when connecting through proxies or firewalls. #971
• Results: Fixed “below” result location rearranging entire editor layout when using split editors. #973
• Elasticsearch/OpenSearch: Fixed introspection failing with “HTTP line is larger than 4096 bytes” on clusters with many indices. #975

Changed

• AI: Added execution plan analysis.
• AI: Added BOYK/local LLM support. Configure any OpenAI-compatible endpoint in settings under dbcode.ai.customModel. #816
• Results: Added setting to control zebra striping in the grid.
• MCP: Added configurable OAuth token expiry (dbcode.ai.mcp.tokenExpiry), default 1 day, max 90 days. #960

Fixed

• MySQL: Fixed MariaDB derivative detection failing when @@GLOBAL.version_comment doesn’t contain “mariadb” (e.g., Debian packages). #968
• SQLite: Fixed truncate using unsupported TRUNCATE TABLE syntax, now uses DELETE FROM. #965
• ClickHouse: Fixed query cancellation not killing the running query on the server. #962
• PostgreSQL: Fixed backup/restore to use local tools first if on the path. #961

Changed

• Missing WHERE Detection: Warns before executing DELETE or UPDATE statements missing a WHERE clause. Configurable per role (allow/ask/deny). Available once per session for Core users, continuously for Pro.

Fixed

• Results: Fixed right-click Filter By/Exclude not working for dates, booleans, and null values in the results grid. Also fixed context menu filters not triggering server-side queries in paging mode. #957
• Results: Fixed Exclude filter to include filtering for null values. #957
• Snowflake: Fixed introspection failing due to unquoted INCREMENT reserved keyword in sequences query. #956
• Snowflake: Fixed regular tables not appearing in the tree when hybrid, transient, or dynamic tables exist.
• Execution: Fixed standalone BEGIN/END blocks in SQL Server and Oracle grouping subsequent statements into one block. #958