This addon uses an external AI service (like OpenAI or MS Azure OpenAI) to help technical documentation writers generate content.

Release Notes 4.0.0

• Security Enhancements - Various security enhancements were implemented. Information regarding favorite prompts, chat history, or AI XPath functions are stored encrypted. The Retrieval-Augmented Generation (RAG) preferences page has some new security-related options to choose whether or not to enable reading and writing content in the project.
• Intuitive Identification of XML Code in Markdown - If the AI returns a response with Markdown content that contains XML code (for example, from a chat session where you give the AI a task such as converting a list into a table), the AI might start with some explanations followed by the actual response wrapped in the ```xml marker. This type of response is now detected and the block of XML code can be inserted or previewed.
• AI Positron Fix with RAG - The AI Positron Fix action now uses Retrieval-Augmented Generation to get more context from the current document.
• Oxygen Feedback based Retrieval Augmentation - If the Oxygen Feedback product is used to provide search for a web site generated from the DITA XML project, the Feedback search system can be used to retrieve related content instead of the current indexed DITA XML project. In the Feedback administrator Installation page the embed HTML information contains an unique token which can be copied and pasted in the AI Positron Retrieval-Augmented Generation (RAG) preferences page in the WebHelp Feedback token field.
• Document type context in chat - If there is content selected in the main editor area when a chat is initiated, the selection is passed to the AI along with the document type as context for the conversation.
• Split Large DITA Topics - When working with DITA XML documents, the new Split Topic AI action (located in the Intelligent Agents category) can be used to split large topics with numerous subjects into multiple topic files. The newly created topics are also conveniently referenced in the DITA map that is currently open in the DITA Maps Manager.
• Expand DITA Topic Draft - The new Expand Draft AI action (located in the Intelligent Agents category) enhances a draft of a DITA XML topic by refining the content, adding markup that is consistent with similar topics, and generating documentation based on the referenced images using Vision support. The AI looks for similar content in topics gathered from the related links or through the Retrieval-Augmented Generation (RAG) process that looks in both the current project and WebHelp output with the Oxygen Feedback component deployed for search functionality. This action is available for DITA XML documents.
• Create DITA Topics - The new Create Topics AI action (located in the Intelligent Agents category) generates DITA XML topics based upon user input, incorporating relevant project content. It creates a topic hierarchy, assigns DITA document types, and proposes a DITA map location. Topics are then saved and added upon user approval. The AI looks for similar content in topics gathered from the related links or through the Retrieval-Augmented Generation (RAG) process that looks in both the current project and WebHelp output (with the Oxygen Feedback component deployed for search functionality). This action is available for DITA XML documents.
• Reuse Component - Added in the 'Reuse' category a new 'Component' action which asks the AI to identify the closest existing reusable component which matches the current selected paragraph.
• Reuse Product Names - Added in the 'Reuse' category a new 'Product Names' action which finds product names in the selected content and replaces them with key references.
• Suggest Refactoring for Selected Code - The new Suggest Refactoring AI action (located in the Development category) can be used to generate a suggestion for refactoring the selected code to simplify it and make it easier to read and understand. This action is available for XSLT and XSD documents.
• New functions:
  • add_to_toc - Modifies a the DITA map and adds a topic reference to it at a certain relative location.
  • find_similar_reusable_content - Retrieves a list of reusable DITA XML components which match keywords provided by the AI.
  • get_content_for_document_url - Retrieves the content of a certain document.
  • get_corresponding_dita_keyrefs - Finds DITA key references that correspond to product names in the context DITA map.
  • get_current_editor_file_location - Retrieves the location (absolute path/URL) of the current file that is open in the editor.
  • get_related_content_from_webhelp - Returns content from the associated Oxygen Feedback WebHelp system for which a token has been configured in the Retrieval-Augmented Generation (RAG) preferences page.
  • get_topic_context_in_toc - Returns from the DITA map opened in the DITA Maps Manager view the hierarchical structure path which references the selected topic.
  • resolve_dita_key_or_content_reference - Resolves a DITA XML key reference, a content key reference, or a content reference to the target text.
  • save_document - Saves the document at a specified URL path and writes content to it.
• New custom action types:
  • prompt-user - Prompt the user for extra details and then perform the action.
  • replace-selection-based-on-image-analysis - Use details obtained from the referenced DITA XML images to perform the action prompt.
• Current Element Highlighted When Applying an AI Action in Text Mode - When utilizing AI actions while you are editing in Text mode, it is now easier to identify the element that is affected by an AI action. This is especially beneficial when there is no specific selection in the editor. The current element, which serves as the context for applying the action, is now highlighted for improved visibility and ease of identification.

Release Notes 3.0.0

• Actions using Retrieval-Augmented Generation (RAG) - For Oxygen version 26.1 (latest build) and newer, certain actions that generate content, such as New DITA Topic or Add Structured Content, now use information from the project opened in the Project view to enrich the AI context and receive more meaningful and project-targeted responses. The new Retrieval-Augmented Generation (RAG) preferences page allows changing various settings related to RAG.
• Chat using RAG - RAG is also enabled for chat sessions in Oxygen version 26.1 (latest build) and newer.
• Generate Documentation Draft using RAG - When the use-related-content-from-project="true" attribute is set on the <draft-summary> element, RAG is enabled for the Generate Documentation Draft action.
• Function Calls - Both built-in and user-defined actions can refer in their definition to existing functions that are to be called by the AI engine when giving a response. This gives actions more context information and allows for more tailored and accurate answers to be constructed.
• AI Positron Assistant Widget - For documents opened in the Author visual editor, you can click the AI Positron Assistant widget located in the top-right corner of the editing area or use the configurable "Ctrl+Alt+Enter" (Windows/Linux) or "Ctrl+Enter" (Mac) shortcut to access useful AI actions.
• Rewrite content - Added a new action that opens a floating input box at the text cursor location, where you can provide the AI with instructions on how to rewrite the selected content (or the current paragraph if there is no current selection). This action is available in the AI Positron Assistant Widget or by using the configurable "Ctrl+Alt+R" (Windows/Linux) / "Cmd+Alt+R" (Mac) shortcut.
• Select AI Model - A drop-down was added to the AI Positron Assistant side view in order to quickly select the default AI model to use for chat and actions.
• Content Generation > Add Structured Content - A new action is available that can be used to replace the current selection with new structured content that is generated based upon similar content from the current project (if available).
• Content Generation > New DITA Topic - For Oxygen version 26.1 (latest build) and newer, the New DITA Topic action now uses related information from the current project when constructing the content and structure of the new topic.
• Rewrite > Improve Structure - A new action is available that can be used to improve the selected DITA XML content by adding additional structure or inline elements.
• Review > Proofread - A new action is available that can be used to add comments in content that has logical consistency problems, grammar or spelling mistakes, or is hard to read and comprehend.
• Review > Resolve Comments - A new action is available that changes the selected content based on suggestions found within comments and then removes the comments.
• Regenerate Response - After you receive a response from the AI, you have the possibility to ask it to generate another one. You can also decide which AI model to be used when generating the new response.
• Default model - For the OpenAI connector, GPT-4o, the most intelligent model from OpenAI, is now the default model in the AI Positron Assistant.
• GPT-4o mini - For the OpenAI connector, the most affordable intelligent model from OpenAI is now available in the AI Positron Assistant.
• Preview Response for Text Mode - When using AI actions while working with your documents in the Text mode, you now have the ability to preview the AI response before inserting the changes in the document.
• For the OpenAI connector, each answer from the AI presented in the chat panel now specifies the model that has been used for generating it.
• The actions related to an AI-generated message in the chat panel (Insert, Replace, Preview, etc.) are now buttons with icons instead of links.

Release Notes 2.1.0

• The new Generate Image Alternate Text action allows generating an alternate text for a DITA XML image that is selected in the main editing area in the Author visual editing mode.
• If you start a conversation with the AI and there is selection in the current editor area, that selection is passed as a context to the chat engine.
• The add-on contributes three new XML refactoring actions in the Oxygen XML Refactoring dialog box, in the new AI category:
  • Generate alternate text for images in DITA XML topics
  • Generate missing short descriptions in DITA XML topics
  • Shorten existing short descriptions in DITA XML topics
• For the OpenAI connector, gpt-4o, the latest model from OpenAI, has been added to the list of supported models.
• The MS Azure connector now supports authentication using Microsoft Entra ID, enhancing security and flexibility.
• When selecting an AI assistant response in the chat view, you now have contextual actions available to insert the selection into the document or preview the insertion.

Release Notes 2.0.0

• Added support to connect to the Anthropic Claude AI engine.
• Added support in the MS Azure OpenAI connector to configure separate Vision Endpoint, Deployment and API Key settings used for analyzing messages which contain images.
• A new action called "Generate Documentation Draft" is available, and it can be used to draft a DITA topic based on a configuration file that fine tunes the generation process.
• A new action called "Update content based on images" is available, and it can be used to update the content of a DITA XML topic based on the images that it refers.
• Images, XML files, and text files can be attached to messages in the Chat panel.
• Useful AI actions (Correct Grammar, Improve Readability, Use Active Voice) are now available as quick assist fixes in the editor by clicking the bulb located on the navigation vertical stripe.
• Two new variables may be used when you write prompts: ${selection-original}, which expands to the selected text with all track changes rejected, and ${selection-final}, which expands to the selected text with all track changes accepted.
• You can now specify in the "Oxygen AI Positron Assistant Enterprise" preferences page which actions to exclude from being presented in the available actions list.
• For the OpenAI connector, the gpt-4-turbo-preview model has been replaced with the recently released gpt-4-turbo.

Release Notes 1.2.0

• Oxygen XML Enterprise edition license now includes access to Oxygen AI Positron Assistant Enterprise at no extra cost.

Release Notes 1.1.1

• Fixed an issue where the connection to MS Azure OpenAI failed when the configured endpoint URL address included a path.

Release Notes 1.1.0

• The new Oxygen AI Positron Assistant Enterprise addon enables direct access to your own OpenAI account or AI service hosted on Microsoft Azure. This can be configured in the "Preferences->Oxygen AI Positron Assistant / AI Service Configuration" page. A special license key is required for this feature.
• The AI Positron Assistant Enterprise side view supports content generation and refinement of the AI response. This includes:
  • 'Content Generation' for creating new DITA XML topics, continuing writing, generating short descriptions, index terms, and following instructions.
  • 'Rewrite' actions for grammar correction, readability improvement, content conversion to active voice, itemization, and joining items.
  • 'Overview' actions for answering and generating questions, summarizing content, and improving readability.
  • 'Translation' actions support English, German, French, and Japanese languages, with an option to provide a target language.
  • 'Marketing' features for creating release notes, marketing posts, improving SEO, and rewriting content using the Pain-Agitate-Solution and Features-Advantages-Benefits frameworks.
• The AI Positron Refactoring action is available for refactoring multiple files simultaneously using AI.
• The 'Validation Quick Fixes' feature allows the use of the AI Positron Fix action to rectify validation issues.
• Custom actions and prompts can be created in the AI Positron Assistant by recording changes.
• Two XPath extension functions are available in the content completion proposals for Schematron, XSLT, XQuery, and XPath:
  • 'ai:transform-content(instruction, content)' for automatic content transformation using AI.
  • 'ai:verify-content(instruction, content)' for automatic content validation using AI.