[py] shinychat 0.5.1

Bug fixes

• pydantic is now a hard dependency. (#252)

[py] shinychat 0.5.0

New features

• Chat() now accepts an optional client= parameter. When provided, streaming, cancellation, bookmarking, and greeting handling are wired up automatically — no manual plumbing required. The chat.client property exposes a ChatClient wrapper with .value (the raw chatlas client), .set() for swapping models mid-session, and .clear() for resetting the conversation with flexible history management.

• Added slash commands: a typeahead command palette that lets users trigger named shortcuts directly from the chat input. Type / to open the palette, filter by typing, and pick a command with arrow keys or click. Commands can expand into LLM prompts, trigger server-side side effects (clear chat, open a modal, export transcript), or be handled entirely client-side via the cancelable shiny:chat-slash-command DOM event. Register commands with chat.slash_command(), which accepts 0- or 1-argument async handlers; 1-argument handlers receive the text typed after the command name as user_input. The echo parameter controls whether an invocation is recorded as a user message and triggers a loading state. Echoed commands are faithfully restored on bookmark/restore. (#239)

• Added submit_key parameter to chat_ui() and Chat.ui(): "enter" (default, Enter submits) or "enter+modifier" (Ctrl/Cmd+Enter submits, plain Enter inserts a line break). The input remains editable while a response is streaming — only submission is blocked, not typing. (#251)

Improvements

• Bookmarking now correctly restores assistant messages that mix content types within a single message — for example a reply that combines markdown, a raw-HTML widget, and reasoning. Previously, restoring such a message could lose formatting or interactive components.

• All navigating links in assistant messages now open in a new tab to preserve the app's session state. Cross-origin links still show the confirmation dialog; same-origin links open directly. (#238)

Breaking changes

• Removed the deprecated format and token_limits parameters from .messages(), the tokenizer parameter from Chat(), and the .transform_user_input() decorator. These features overreached into LLM provider responsibilities; use your provider (e.g., chatlas, LangChain) to manage conversation formatting, token limits, and input transformation instead. Calling any of these now raises a TypeError with migration guidance. (#245)

Bug fixes

• Fixed the copy button on code blocks not working in some embedded contexts. (@thisisnic, #247)

• Fixed the external link confirmation dialog not rendering in Safari. The backdrop overlay appeared but the dialog content was invisible due to a Bootstrap/<dialog> CSS interaction. (#201, #238)

• Fixed pressing Escape to dismiss the external link dialog leaving it in a broken state where subsequent link clicks no longer worked. (#238)

• Fixed a bug where <thinking> tags inside fenced code blocks (```) or inline backtick spans were incorrectly treated as model reasoning content and hidden from the visible output. Also fixes a chunk-boundary false positive where a chunk ending with a non-newline character followed by a chunk beginning with <thinking> would enter thinking mode. (#235)

[r] shinychat 0.4.0

Experimental internal changes

• The chat UI's rendering layer has been migrated from Lit to React. This significantly improves streaming performance — incoming chunks no longer clear previous DOM state — and makes the codebase more maintainable. One trade-off is that certain Shiny UI elements embedded in chat messages may not work as well as before (e.g., inline <script> tags are generally not supported inside a React runtime). If you encounter issues, please let us know.

New features and improvements

• The chat UI now displays model reasoning/thinking content as collapsible panels above assistant responses. Thinking content streams in real-time with animated topic labels. This works with providers that support structured thinking (e.g., Claude's extended thinking via ellmer) and with local models that wrap reasoning in <thinking> tags. (#208)

• Added enable_cancel parameter to chat_ui() to show a stop button that lets users cancel an in-progress AI response. Press the stop button or hit Escape to cancel. chat_mod_ui() enables cancellation by default, and chat_mod_server() handles the cancellation wiring automatically, using the stream cancellation features introduced in ellmer v0.4.1. (#221)

• Markdown lists where every item is a <span class="suggestion"> are now rendered as a grid of clickable suggestion cards. Each suggestion's text content becomes both the card label and the value sent on click. To add a short heading above the body text, set the title attribute on the span — e.g. <span class="suggestion" title="Heading">Body text shown on the card.</span>. Only the body text (not the title) is submitted when the card is clicked. Cards stream in with staggered animations and support keyboard navigation (arrow keys, Home/End) with roving tabindex. (#219)

• Added chat_greeting() for creating welcome messages that appear when the chat is empty. Greetings can be set statically via chat_ui(greeting=) or dynamically from the server with chat_set_greeting(). They are automatically dismissed when the user sends their first message. A new greeting_requested input fires when the chat is visible, empty, and has no greeting, enabling LLM-generated welcome messages. chat_mod_server(greeting=) accepts a function for auto-generated greetings. (#217)

• Tool result cards now render images and PDFs returned by ellmer tools. When a tool returns content_image_file(), content_image_url(), or content_pdf_file(), the result is displayed as an inline image or a PDF filename badge. Mixed content lists (e.g., list(ContentText("summary"), content_image_file("plot.png"))) are rendered with items interleaved in order. (#225)

• Added footer parameter to chat_ui() for displaying arbitrary HTML content below the chat input. Useful for disclaimers, attribution, or interactive toolbars. Styled with sensible defaults and customizable via --shiny-chat-footer-font-size and --shiny-chat-footer-color CSS custom properties. (#224)

• Tool result cards now support a fullscreen toggle. Set full_screen = TRUE in the display list (or set res$full_screen <- NA in a custom contents_shinychat() method) to add a button that expands the card to fill the viewport. Press Escape, click the backdrop, or use the close button to exit fullscreen.

• Added footer field to ToolResultDisplay for displaying custom HTML content below the tool result card body. (#178)

• chat_mod_server() now returns a set_client(new_client, sync = TRUE) function for swapping the chat client used by the module at runtime. When sync = TRUE (the default), the new client inherits the current conversation's turns, system prompt, and tools so the conversation continues seamlessly. If a response is currently streaming, the swap is deferred until the stream completes. (#227)

• chat_mod_server() now returns a status reactive that reports the current interaction state: "idle" when no response is in progress, or "streaming" while a response is actively being received. (#227)

• chat_restore() now invisibly returns a cancel function that tears down all bookmark registrations made by that call. This is useful when swapping the chat client via set_client(), which handles the re-registration automatically. (#227)

Improvements

• All navigating links in assistant messages now open in a new tab to preserve the app's session state. Cross-origin links still show the confirmation dialog; same-origin links open directly. (#238)

Bug fixes

• Fixed the external link confirmation dialog not rendering in Safari. The backdrop overlay appeared but the dialog content was invisible due to a Bootstrap/<dialog> CSS interaction. (#201, #238)

• Fixed pressing Escape to dismiss the external link dialog leaving it in a broken state where subsequent link clicks no longer worked. (#238)

• Fixed an issue where user chat messages would display the default assistant icon. (#162)

[py] shinychat 0.4.0

New features

• The chat UI now displays model reasoning/thinking content as collapsible panels above assistant responses. Thinking content streams in real-time with animated topic labels. This works with providers that support structured thinking (e.g., Claude's extended thinking via chatlas) and with local models that wrap reasoning in <thinking> tags. (#208)

• Added chat_greeting() for creating welcome messages that appear when the chat is empty. Greetings can be set statically via chat_ui(greeting=) or dynamically with Chat.set_greeting(), and are dismissed when the user sends their first message. A new {id}_greeting_requested input fires when the chat is visible, empty, and has no greeting, enabling LLM-generated welcome messages. (#217)

• Added enable_cancel parameter to chat_ui() and Chat.ui() to show a stop button that lets users cancel an in-progress AI response. Press the stop button or hit Escape to cancel. Wire up the input.<id>_cancel event to a chatlas.StreamController (introduced in chatlas v0.18.0) to connect the UI to your chat provider. (#221)

• Added footer parameter to chat_ui() and Chat.ui() for displaying arbitrary HTML content below the chat input. Useful for disclaimers, attribution, or interactive toolbars. Styled with sensible defaults and customizable via --shiny-chat-footer-font-size and --shiny-chat-footer-color CSS custom properties. (#224)

• The chat input now supports history navigation: press Up/Down arrow keys when the input is empty to cycle through previously sent messages. Editing a recalled message locks navigation until the input is cleared. (#222)

Improvements

• Tool result cards now render images and PDFs returned by chatlas tools. When a tool returns ContentImageInline, ContentImageRemote, or ContentPDF, the result is displayed as an inline image or a PDF filename badge. Mixed content lists (e.g., [ContentText("summary"), content_image_file("plot.png")]) are rendered with items interleaved in order. Standalone image and PDF content items in turn history are also rendered correctly. (#225)

• Markdown lists where every item is a <span class="suggestion"> are now rendered as a grid of clickable suggestion cards. Each suggestion's text content becomes both the card label and the value sent on click. To add a short heading above the body text, set the title attribute on the span — e.g. <span class="suggestion" title="Heading">Body text shown on the card.</span>. Only the body text (not the title) is submitted when the card is clicked. Cards stream in with staggered animations and support keyboard navigation (arrow keys, Home/End) with roving tabindex. (#219)

• Updated minimum chatlas version to >=0.15.0. (#208)

Bug fixes

• Fixed a circular import error triggered when import shinychat ran before shiny had fully initialized. (#212)

• Fixed extra newlines appearing when copying user message text. (#209)

[py] shinychat 0.3.2

Bug fixes

• Updated for compatibility with htmltools 0.7.0. split_html_islands now recognizes the new TagifiedTag / TagifiedTagList sibling classes when deciding which children carry the data-shinychat-react attribute, and Tool*Component.tagify() implementations are annotated with htmltools.Tagified and return fully-tagified output per the new Tagifiable contract. (#226)

[py] shinychat 0.3.1

Bug fixes

• Fixed an issue where Shiny UI components (e.g., inputs) passed to MarkdownStream's content parameter could fail to initialize, especially on WebKit-based browsers. (#205)

[py] shinychat 0.3.0

Experimental internal changes

• The chat UI's rendering layer has been migrated from Lit to React. This significantly improves streaming performance — incoming chunks no longer clear previous DOM state — and makes the codebase more maintainable. One trade-off is that certain Shiny UI elements embedded in chat messages may not work as well as before (e.g., inline <script> tags are generally not supported inside a React runtime). If you encounter issues, please let us know.

New features

• Tool result cards now support a fullscreen toggle. Set full_screen=True in ToolResultDisplay() to add a button that expands the card to fill the viewport. Press Escape, click the backdrop, or use the close button to exit fullscreen. (#179)

• Added footer parameter to ToolResultDisplay for displaying custom HTML content below the tool result card body. (#178)

Breaking changes

• Removed the deprecated transform_user and transform_assistant parameters from .messages(). As a result, .messages() now always returns transformed content (i.e., the result of applying .transform_user_input() / .transform_assistant_response(), if any), meaning it's better reflection of UI state than LLM-facing message state. This change reflects a greater change in philosophy that shinychat shouldn't be managing LLM message state (a backend framework like chatlas, langchain, pydantic, etc. should do this instead). (#193)

• Removed the deprecated transform parameter from .user_input(). .user_input() now always returns the currently stored user message content, which matches the simplified transform handling used by .messages(). (#193)

• ChatMessageDict (returned by .messages()) may now include an html_deps key containing serialized HTMLDependency dicts. Code that unpacks or iterates these dicts with a fixed set of keys should be updated to handle the new field. (#193)

Improvements

• Streaming messages now support mid-stream content type transitions. When the content type changes during streaming, previous content is preserved as a frozen segment, maintaining HTML islands and Shiny bindings. (#199)

• Replaced the custom auto-scroll implementation with the use-stick-to-bottom library for smoother stick-to-bottom behavior during streaming. A centered "scroll to bottom" button now appears when the user scrolls away from the bottom. (#195)

• Migrated Google provider from the deprecated google-generativeai SDK to google-genai. (#174)

Bug fixes

• Fixed some issues with bookmarking. (#188, #192, #193)

• Fixed chat message content touching the right edge of the container. (#197)

• Fixed content inside tool cards (e.g., widgets, plots) not re-laying out when the card is collapsed or expanded. (#180)

• Fixed the chat UI crashing on non-HTTPS contexts (e.g., RStudio Server over plain HTTP) due to crypto.randomUUID() being unavailable outside secure contexts. (#186, #187)

• Fixed pyright 1.1.409 compatibility by adding missing __all__ exports. (#200)

[py] shinychat 0.2.9

Improvements

• Improved scroll to bottom behavior with tool requests/results. (#157)
• Constrain images in markdown content with max-width: 100%. (#168)

Maintenance

• Fixed an issue where user chat messages would display the default assistant icon. (#162)
• shinychat now requires Python 3.10 or later and, optionally, langchain 1.0.0 or newer. (#156)

r-shinychat 0.3.0

Breaking changes

• chat_mod_server() now returns a list of reactives for last_input and last_turn, as well functions to update_user_input(), append() and clear() the chat. (#130, #143, #145)

New features

• Added chat_restore() which adds Shiny bookmarking hooks to save and restore the {ellmer} chat client. (#28, #82)

• Added update_chat_user_input() for programmatically updating the user input of a chat UI element. (#78)

• shinychat now shows tool call request and results in the UI, and the feature is enabled by default in chat_app() and the chat module (chat_mod_server()). When using chat_append() with chat_ui(), set stream = "content" when you call the $stream_async() method on the ellmer::Chat client to ensure tool calls are included in the chat stream output. Learn more in the tool calling UI article. (#52)

• Added chat_append(icon=...) and chat_ui(icon_assistant=...) for customizing the icon that appears next to assistant responses. (#88)

Improvements

• chat_app() now correctly restores the chat client state when refreshing the app, e.g. by reloading the page. (#71)

• External links in chat messages in chat_ui() now open in a new tab by default, with a confirmation dialog. (#120)

Bug fixes

• The chat input no longer submits incomplete text when the user has activated IME completions (e.g. while typing in Japanese or Chinese). (#85)

Internal changes

• We consolidated the <shiny-chat-message> and <shiny-user-message> components into a single <shiny-chat-message> component with a data-role attribute to indicate whether it's an "assistant" or "user" message. This likely has minimal impact on your apps, other than custom styles. You should update any shiny-user-message rules to use shiny-chat-message[data-role="user"]. (#101)

• The chat UI's send input button is now identified by the class .shiny-chat-btn-send. (@DeepanshKhurana, #138)

py-shinychat 0.2.8

Bug fixes

• Allow chatlas.types.ToolResultDisplay to be imported without pydantic.