8.1.3 26.0+ plugin Syncro Soft Oxygen AI Positron Oxygen AI Positron

Oxygen AI Positron

Oxygen AI Positron for Desktop is an advanced AI-powered solution designed to assist content creators, technical writers, and developers throughout every stage of the writing process. By leveraging the capabilities of AI platforms, AI Positron streamlines content creation and enhances productivity, thus empowering users to craft high-quality documentation and content with greater speed and confidence.

Release Notes 8.1.3

  1. Fixed an issue where the AI Positron side view remained un-initialized after Oxygen started.

Release Notes 8.1.2

  1. Fixed an issue where continuing the conversation in chat was not working correctly when data was filtered using the Sensitive Data Masking support.
  2. Improved security by using stronger encryption algorithms for options stored encrypted in Positron.
  3. New Models Available - Claude Sonnet 4.6 and Claude Opus 4.6 are now available in the list of models to choose from.

Release Notes 8.1.1

  1. Fixed an issue where stopping the chat when an unsafe tool call (MCP for example) was waiting for user approval allowed the tool to be executed.
  2. Fixed an issue where table specific selections (e.g., selecting multiple table cells or columns) were no longer working in Oxygen XML after AI Positron was installed.

Release Notes 8.1.0

  1. Sensitive Data Masking for AI Requests - Automatically identifies and masks personal information (emails, phone numbers, API keys) in AI requests, with support for adding custom patterns to extend data protection.
  2. Improved Review Experience for AI-Modified Files - New collapsible file list in the chat area allows for direct review of AI changes; users can accept or discard all modifications at once or process specific files individually.
  3. New Default Models - Claude Haiku 4.5 is the new default for the Positron Service and GPT-5.1 is the new default for the OpenAI connector.
  4. OpenAI GPT-5.2 Models Available - The flagship GPT-5.2 and GPT-5.2 Thinking models are now available in the list of OpenAI models to choose from.
  5. More Context Information Available to Chat Modes - When chatting with the AI additional context information is automatically made available to the LLM:
    • Name and version of the Oxygen application.
    • Operating system locale, name and version.
    • Current date.
    • Project directory and main files information.
    • Location of DITA Map currently opened in the DITA Maps Manager view.
    • Location of current edited document.
    • Information about caret and selection intervals in the current edited document (Oxygen 28.1 or newer).
  6. Chat Modes Prompt and Tools Improvements - Improved the prompts and tools available to all chat modes.
  7. New Chat Mode: Planner - A new 'Planner' chat mode creates a comprehensive plan by exploring your project, then invokes the right agents to execute each part of the plan. This new chat mode can be used for heavy lifting, for making lots of project wide changes without running into LLM context limitations.
  8. Smarter Attachment Expansion in Planner Mode - Automatically parses PDF and Word (.docx) attachments into Markdown sections and extracts images, enabling Planner subagents to process, transform, or convert document components independently.
  9. Tools Selector - The 'Tools' selector button next to the 'Chat Modes' selector allows you to disable various tools (either added via 'Model Context Protocol' or provided in the chat mode definition) for the current selected chat mode.
  10. Tool to run command line - All chat modes now have access to the tool "run_command_line" which can be used to invoke a command line process. The tool needs explicit user approval when it is invoked.
  11. Tools to list and invoke AI agents - With the newly added tools 'retrieve_ai_agents' and 'invoke_ai_agent' new powerful chat modes can be created to plan complex changes and then use separate agents for each part of the plan.
  12. Model Context Protocol update - Updated the used MCP SDK to version 0.16.0.
  13. Retrieve range of lines from document - The tool which retrieves content from a document can now be used by the LLM to retrieve a range of lines from the document.
  14. Tool to Copy, move or rename a file/folder - All chat modes now have access to the tool "copy_move_rename_path" which can be used to copy, move or rename a file or a folder. The tool needs explicit user approval when it is invoked.
  15. Chat Selection Includes Images - When using the Chat view, if the selection in the document contains images they are sent to the LLM along with the selected text.
  16. Create New Chat Modes - You can now officially define new chat modes by creating new AI actions with the "output-action": "agentic" property.
  17. Control AI Action Visibility - You can now control if the AI action you define is visible in the UI or only to other AI actions as a sub-agent using the new the "visibility" property.
  18. Create AI Actions with Vision Support - You can now control if a custom defined AI action sends to the LLM both the content and the referenced images using the "input-type":"markup-with-images" JSON property.
  19. New DITA Topic Prompt Improvements - The prompt for the predefined 'New DITA Topic' AI action was improved to create better results.
  20. Reuse Component Prompt Improvements - The prompt for the predefined 'Reuse Component' AI action was improved to create better results.
  21. Generate Image Alternate Text Improvements - The alternate text generation for images now works on multiple images referenced in a document.
  22. Preserve Structure and Styles for Attached Word and PDF Documents - When attaching Word (docx) and PDF documents in the Chat view their internal structure (headings, lists, tables, code blocks) and styling (bold, italic) are better preserved.
  23. Removed support for older models - Support for older models, such as GPT-4, GPT-4 Turbo, GPT-4o, Sonnet 3.7, and others, has been removed.