Release v1.4.0

What's Changed

• Prepare 1.4.0 release by @rhennigan in #524
• Release v1.4.0 by @rhennigan in #525

Full Changelog: v1.3.9...v1.4.0

Release v1.3.9

What's Changed

• Bugfix: "ChatCodeInlineTemplate" template boxes cause failures when exploding cells by @rhennigan in #522
• Release v1.3.9 by @rhennigan in #523

Full Changelog: v1.3.8...v1.3.9

Release v1.3.8

What's Changed

• Bugfix: Fix for inconsistent EvaluationCell behavior on Linux and Mac by @rhennigan in #518
• Release v1.3.8 by @rhennigan in #519

Full Changelog: v1.3.7...v1.3.8

Release v1.3.7

What's Changed

• Add workaround for EvaluationCell[] failures in AttachCodeButtons by @rhennigan in #511
• Bugfix: Exploding cells failure when there's a language-agnostic code block by @rhennigan in #512
• Release v1.3.7 by @rhennigan in #513

Full Changelog: v1.3.6...v1.3.7

Release v1.3.6

What's Changed

• Bugfix: Workaround for multi-line code blocks not parsing correctly in cloud notebooks by @rhennigan in #502
• Bugfix: Fix the missing disable button when auto assistance outputs are first created by @rhennigan in #503
• Show checkmark next to "User" role as default by @rhennigan in #504
• Bugfix: Return Null instead of generating an internal failure if the current chat input is excluded by @rhennigan in #505
• Bugfix: Avoid messages that occur during bad caching of tool response boxes by @rhennigan in #507
• Feature: MaxContextTokens can be set to Infinity to avoid truncation by @rhennigan in #508
• Release v1.3.6 by @rhennigan in #509

Full Changelog: v1.3.5...v1.3.6

Release v1.3.5

What's Changed

• LLMServices UI integration by @rhennigan in #494
• Bugfix: Add scrollbars to submenus when large to avoid content clipping by @rhennigan in #497
• Release v1.3.5 by @rhennigan in #498

Full Changelog: v1.3.4...v1.3.5

Release v1.3.4

What's Changed

• Bugfix: Fix internal failure for custom tools in 1.3.3 by @rhennigan in #493
• Release v1.3.4 by @rhennigan in #495

Full Changelog: v1.3.3...v1.3.4

Release v1.3.3

What's Changed

• Bugfix: Better error handling for outer sandbox evaluator timeouts by @rhennigan in #480
• Added a WolframLanguageEvaluator tool for cloud by @rhennigan in #481
• Avoid serializing giant tokenizers into tagging rules of chat outputs by @rhennigan in #482
• Bugfix: Fix tokenizerName["gpt-4-vision-preview"] incorrectly giving "gpt-4" instead of "gpt-4-vision" by @rhennigan in #483
• Make inlineInteractiveCodeCell portable between cloud and desktop and avoid bouncing on mouseover in cloud by @rhennigan in #484
• Bugfix: dynamic box was preventing editing and code buttons were failing by @rhennigan in #486
• Use different URI schemes to inform the LLM about interactivity by @rhennigan in #487
• Release v1.3.3 by @rhennigan in #488

Full Changelog: v1.3.2...v1.3.3

Release v1.3.2

What's Changed

• Test OpenAI paclet for multimodal support by @rhennigan in #469
• Misc multimodal bug fixes by @rhennigan in #470
• Encourage the LLM to fetch relevant pages in web search results by @rhennigan in #471
• Include commit info in builds created for layout by @rhennigan in #472
• Print debug info to build log if there's a failure by @rhennigan in #474
• Bugfix: Do not issue messages for tools that have been removed but still have lingering selection settings by @rhennigan in #476
• Bugfix: Handle failures in tool call chat submit by @rhennigan in #477
• Release v1.3.2 by @rhennigan in #478

Full Changelog: v1.3.1...v1.3.2

Release v1.3.1

What's Changed

• Support step-by-step solutions in WA output by @rhennigan in #430
• Bugfix: incomplete definition for truncateString by @rhennigan in #433
• Custom Processing Functions by @rhennigan in #435
• Tell Wolfie his name by @rhennigan in #437
• Fixed an issue with processing responses for PaLM by @rhennigan in #438
• Support "Prompts" property in LLMEvaluator association specs by @rhennigan in #439
• Minor QoL updates by @rhennigan in #440
• WTC 2023 Notebooks by @rhennigan in #441
• Support for markdown tables in input and output by @rhennigan in #444
• Added highlighting of included cells when mousing over chat inputs by @rhennigan in #445
• Add support for GPT-4-Vision and longer contexts by @rhennigan in #447
• Improved error handing in web search and web image search tools by @rhennigan in #448
• Improved serialization of definition notebook structures by @rhennigan in #449
• Bugfix: Explode cells wasn't handling section cells properly by @rhennigan in #450
• Bugfix: Serialize multimodal messages in feedback submissions by @rhennigan in #451
• Bugfix: Cut message content when exceeding token budget by @rhennigan in #453
• Disable tools for PaLM by default by @rhennigan in #452
• Inline WL code input boxes by @rhennigan in #454
• Multimodal tool responses by @rhennigan in #455
• Minor development improvements by @rhennigan in #457
• Update license and add contributor notice by @rhennigan in #458
• feat: improve setting ui in UI.wl, add OpenAi Chat Completion URL settings by @nhannht in #456
• Customizable tool call frequency by @rhennigan in #460
• Bugfix: Cloud notebooks use two arguments for BoxData by @rhennigan in #462
• Workaround for cloud current value inheritance by @rhennigan in #463
• Temporary workaround for ChatObject not supporting multimodal content by @rhennigan in #464
• Workaround for FrontEndResource["ContextMenus", ...] failing when used within docker by @rhennigan in #466
• Bugfix: Fix missing tool call examples from tool prompting by @rhennigan in #467
• Release v1.3.1 by @rhennigan in #468

New Contributors

• @nhannht made their first contribution in #456

Full Changelog: v1.3.0...v1.3.1