Release Notes

All notable changes to ProActions are documented in this file.

Recent Highlights

What's New in 1.2.0 (upcoming)

📊 Flow Monitor - Real-time visualization and control of running workflows
⏱️ Progress Tracking - Percentage, step-based, and indeterminate progress indicators
🛑 Cancellation Support - Stop long-running flows with confirmation dialogs
💬 Interactive Feedback - Prompts and actions within the monitor
🔄 Streaming Content - Real-time LLM response display as it arrives
🎨 Smart Auto-Hide - Intelligent visibility based on execution state
📄 Document Lifecycle Steps - New SAVE_DOCUMENT and CLOSE_DOCUMENT steps with cross-platform support in Swing and Prime
⚡ Event-Driven Actions - New internal event system with action subscriptions and EMIT_EVENT / DISPATCH_EVENT for trigger-based automation
🤖 Agent Skills - Load reusable skill instructions from the server into AI completion steps
🛠️ Built-in AI Tools - Ready-made tools for document access, user interaction, and metadata; no custom function definitions required
🔌 MCP Integration - Call MCP server tools directly from flows with HUB_MCP_TOOLS and HUB_MCP_INVOKE steps
🏗️ Block Drop Integration - Trigger actions automatically when blocks are dropped in the Swing editor (onBlockDrop)
🧩 New Control & Utility Steps - PLATFORM, GET_ZONES, GET_PURE_TEXT, TAG_PURE_TEXT, CLEAR_TAGS
⚙️ Section-Level Configuration - Override monitor and other settings at section level

What's New in 1.1.0 (latest)

🎨 Interactive Diff - Review and selectively accept/reject text changes with granular control
⚡ Inline Steps - Improved UX with background processing for FORM, USER_SELECT, and SHOW_RESPONSE
🔧 Swing Integrations - Native Swing toolbar, object action, and context menu support
🤖 Enhanced Chat Completion - Tool calling and improved multimodal support
🔄 New Control Steps - BREAK, FOR, TRY, and PARALLEL for advanced workflow logic
🎵 Audio Playback - New PLAY_AUDIO step with customizable player
📚 Prompt Library - Reuse stored prompts from Swing Prompt Management UI

Major Features Since 1.0.0

Flow Monitor (1.2.0) - Real-time workflow visualization and control
Agent Skills (1.2.0) - Reusable AI instruction sets loaded from the server
Built-in AI Tools (1.2.0) - Ready-made tools for AI function calling
MCP Integration (1.2.0) - Call MCP tools directly from flows or expose to AI
Block Drop Integration (1.2.0) - Trigger flows on drag-and-drop in Swing editor
Interactive Diff (1.1.0) - Granular change review and selection
Swing Integration (1.1.0) - Native inline toolbar and action integration
Schema Support (1.0.11) - Auto-completion and validation for YAML configs
Form Enhancements (1.0.10) - Rich form components (HTML, Markdown, Diff)
Progress Bars (1.0.8) - Visual feedback for long-running workflows
Context Menus (1.0.7) - Text selection and one-click menus
ProActions Hub - Centralized API key management for 40+ AI providers
Multi-Language Support - English, Spanish, French, Italian, German

Version History

1.2.0 - March, 2026 - Flow Monitor, Agent Skills, Built-in AI Tools, MCP Integration, Block Drop
1.1.0 - October 24, 2025 - Interactive Diff, Swing Integration, Enhanced Workflows
1.0.12 - February 7, 2025 - Slash Menu, ElevenLabs, Prime 8
1.0.11 - April 28, 2025 - Schema Support, YouTube Integration
1.0.10 - February 27, 2025 - Form Enhancements, Spanish Language
1.0.9 - January 26, 2025 - DeepL Write Integration
1.0.8 - December 13, 2024 - Progress Bars, Dashboard Actions
1.0.7 - November 14, 2024 - Context Menus, External Config
1.0.6 - October 11, 2024 - Permissions, DEBUG Step
1.0.5 - September 3, 2024 - Enhanced Metadata Access
1.0.4 - August 15, 2024 - Performance Improvements
1.0.3 - August 14, 2024 - Stability AI Integration

Detailed Release Notes

1.2.0 - March 2026

📊 Flow Monitor

A comprehensive monitoring system that provides real-time visualization and control of running ProActions workflows:

Real-time Visualization: Monitor displays live progress, current step, and elapsed time for all running flows
Progress Tracking: Multiple progress types supported:
- Percentage-based (0-100%)
- Step-based ("3 of 10 steps")
- Indeterminate (spinner for unknown duration)
Cancellation Support: Users can cancel long-running flows with optional confirmation dialogs
Interactive Feedback: Display prompts, confirmations, and actions directly in the monitor
Streaming Content: Real-time display of LLM streaming responses as they arrive
Smart Auto-Hide: Intelligent hiding based on execution state:
- Never hides on errors (user must acknowledge)
- Never hides with pending interactive prompts
- Never hides during streaming
- Respects user hover and manual close intent
Customization: Full control over position, theme, visibility, and behavior
Permissions: User/group/team-based access control
Multi-App Support: Works seamlessly in both Swing and Prime applications
Keyboard Shortcuts: Navigate flows and dismiss the monitor without leaving the keyboard
Auto-Scroll: Automatically scrolls to the bottom of the monitor log as new messages arrive, with a dedicated scroll-to-bottom button

Configuration:

Global configuration:

AI_KIT:
  MONITOR:
    enabled: true # Enable monitor globally
    position: bottom-right # Position on screen
    theme: dark # Visual theme (dark/light)
    autoHide: true # Auto-hide when flows complete
    autoHideDelay: 3000 # Delay before hiding (ms)
    maxVisible: 5 # Max concurrent executions shown
    apps: # Which apps to show monitor in
      - all # Options: swing, prime, all
    showProgress: true # Show progress bars
    showStepNames: true # Show current step name
    showElapsedTime: true # Show elapsed time
    enableCancellation: true # Allow cancelling flows
    confirmCancellation: true # Confirm before cancelling
    renderMarkdown: true # Render markdown in feedback

Action-level override:

- title: 'Long Running Task'
  monitor:
    enabled: true
    showProgress: true
    confirmCancellation: false
  flow:
    - step: HUB_COMPLETION
      instruction: 'Generate comprehensive report...'

Section-level override:

AI_KIT:
  SECTIONS:
    - section: 'My Section'
      monitor:
        enabled: false # Disable monitor for all actions in this section
      actions:
        - title: 'Quick Action'
          flow:
            - step: SHOW_NOTIFICATION
              message: 'Done'

Use Cases:

Long-running workflows: Provide visibility into multi-step LLM operations
Batch processing: Track progress through large datasets
Interactive flows: Handle user prompts without interrupting execution
Debugging: Understand workflow execution and identify bottlenecks
User training: Demonstrate what ProActions is doing in real-time

See the Flow Monitor Guide for complete documentation.

⚡ Event-Driven Action Triggers

Actions can subscribe to internal events using events in action configuration.
New EMIT_EVENT (DISPATCH_EVENT) step emits custom events for automation chains.
Trigger policies support skip, queue, and parallel behavior while an action is running.
Supported event sources: aikit, system, custom.

See the Swing Integration Guide for full documentation on event-driven triggers.

🤖 Agent Skills

Skills are reusable instruction sets stored on the server that can be dynamically loaded into AI completion steps at runtime. This separates prompt engineering from flow configuration and makes instruction sets reusable across multiple actions.

Skills are stored as SKILL.md files in /SysConfig/ProActions/Skills/<skillName>/
Reference documents (e.g., style guides, terminology lists) can be stored under <skillName>/references/
When loaded, skills are injected into the system prompt automatically
A built-in read_skill_reference tool is exposed to the AI so it can retrieve reference documents on demand
The base path for skills is configurable via skillsBasePath in pro-actions-config.js

- step: HUB_COMPLETION
  behavior: 'You are a writing assistant.'
  instruction: 'Improve this article: {textContent}'
  skills:
    - editorial-style    # Loads /SysConfig/ProActions/Skills/editorial-style/SKILL.md
    - seo-guidelines     # Loads /SysConfig/ProActions/Skills/seo-guidelines/SKILL.md

See the AI Integration Guide — Agent Skills for full documentation including loading modes, reference documents, and the base path configuration.

🛠️ Built-in AI Tools

ProActions now provides a library of pre-configured tools that can be enabled for AI function calling without writing custom function definitions. These tools give the LLM direct access to document content, metadata, and user interaction.

Content tools: getTextContent, getXmlContent, getTextAtXpath, replaceTextAtCursor, insertXmlAtCursor, and many more
Document info tools: getDocumentId, getMetadata, getWorkflowStatus, getChannel, etc.
Container tools: For container (report/DWP) documents
User interaction tools: askSingleChoice, askMultipleChoice, askFreeformQuestion, askConfirmation — pause the AI and prompt the user directly from the monitor
Tool aliases: Define virtual tools with preset arguments to simplify what the AI sees
Alias templates: Share configuration across multiple aliases using aliasTemplate / extend
Auto-discovery: Use autoDiscoverTools: true to let the model discover and activate available tools on demand

- step: HUB_COMPLETION
  instruction: 'Classify the article and update its metadata'
  builtinTools:
    - ContentTools
    - DocumentInfoTools
    - alias: getHeadline
      target: getTextAtXpath
      args:
        xpath: '/doc/story/headline'

See the Built-in AI Tools Reference for the complete tool catalog.

🔌 MCP (Model Context Protocol) Integration

Connect to MCP servers via ProActions Hub and expose their tools to LLMs or invoke them directly from flows.

HUB_MCP_TOOLS: List all available tools from an MCP server
HUB_MCP_INVOKE: Directly invoke an MCP tool from a flow step (without an LLM)
Use builtinTools with mcp: entries to expose MCP tools to an AI completion step

# Invoke an MCP tool directly
- step: HUB_MCP_INVOKE
  server: filesystem
  tool: read_file
  arguments:
    path: '/documents/template.md'

# Expose MCP tools to an LLM
- step: HUB_COMPLETION
  instruction: 'Read and summarize the template'
  builtinTools:
    - mcp: 'filesystem'
      only: ['read_file']

See the AI Integration Guide for full documentation.

🔄 Streaming LLM Responses

OpenAI completion steps now support streaming mode. Tokens from the LLM are displayed in the Flow Monitor in real time as they arrive, providing immediate feedback for long completions.

Streaming works with both plain text and JSON (response_format: 'json_object') responses
Tool call streaming is also supported
Streaming is transparent — the final flowContext.text / flowContext.object output is identical to non-streaming mode
The monitor's auto-hide behaviour is respected during streaming (the monitor stays open until the response is complete)

🏗️ Block Drop Integration (Swing)

Trigger ProActions flows automatically when a block element is dropped in the Swing editor. This enables powerful automation for drag-and-drop scenarios such as cross-story copies, channel migrations, and version restores.

- title: 'Auto-translate Dropped Content'
  hidden: true
  swing:
    onBlockDrop:
      enable: true
      runWhen:
        isCrossStory: true
      when: "options.sourceLanguage !== options.targetLanguage"
  flow:
    - step: DEEPL_TRANSLATE
      instruction: '{{ flowContext._dropInfo.xmlContent | safe }}'
      target_lang: '{{ flowContext._dropInfo.targetLanguage }}'
    - step: REPLACE_XML
      at: CURSOR_PARAGRAPH

Drop information is available via flowContext._dropInfo (blockType, xmlContent, textContent, sourceLanguage, targetLanguage, isCrossStory, etc.).

See the Swing Integration Guide for all configuration options.

🧩 New Steps

PLATFORM: Route flow execution to platform-specific branches (swing, prime, standalone) with an optional default fallback. Enables writing portable flows that behave differently depending on where they run.

- step: PLATFORM
  swing:
    - step: REPLACE_TEXT
      at: CURSOR
  prime:
    - step: SHOW_RESPONSE
      title: 'Result'
  default:
    - step: SHOW_NOTIFICATION
      message: '{{ flowContext.text }}'

GET_ZONES: Retrieve zone and linked document information from a DWP or Report. The result is stored in flowContext.object as an array of zone descriptors.
GET_PURE_TEXT: Extract plain text from the document with XML tags removed and optional character normalization. Useful when you need clean, AI-consumable text without markup.
TAG_PURE_TEXT: Mark segments of text in the document with a virtual or real XML tag. Useful for highlighting AI-identified segments (e.g., claims for fact-checking, passages for review).
CLEAR_TAGS: Remove all virtual tags that were previously added by TAG_PURE_TEXT.

⚙️ Step Improvements

SCRIPTING — returnAs option: Treat the script body as an expression and assign the result directly to a flowContext variable, without needing an explicit return statement. Useful for concise one-liner scripts.
```
- step: SCRIPTING
  script: flowContext.items.filter(i => i.active).length
  returnAs: activeCount
```
SANITIZE — normalizeCharacters: New option to normalize typographic characters (curly quotes, em dashes, etc.) to their ASCII equivalents during sanitization.
SHOW_RESPONSE — Markdown mode: Render the response as formatted Markdown in the modal dialog instead of plain text, enabling richer formatted output.
FORM step: The FORM step now has full schema support, documentation, and examples. All field types and form configuration options are validated.

🔲 Prime Integration Improvements

Significant improvements to the Prime adapter:

isPageContext support: Flows can now detect whether they are running in a page context within Prime, enabling page-aware logic.
Improved XML read/write operations and context detection across a broader set of Prime editor modes.
Better compatibility with the Prime 8 command palette and resizable palette via CHANGE_VIEW_SIZE.

⚠️ BREAKING CHANGE: Client adapter data APIs are now async

To remove blocking synchronous HTTP requests (Prime), the following client methods now need to be executed asynchronously:

getBasefolder(storyId, type)
getUserData(name)
setUserData(name, data)

If you use these methods in SCRIPTING, migrate to async mode and await:

- step: SCRIPTING
  mode: async
  script: |
    const prompts = await client.getUserData('prompts');
    prompts.push({ id: 'new', text: 'example' });
    await client.setUserData('prompts', prompts);
    flowContext.baseFolder = await client.getBasefolder(client.getDocumentId(), 'image');
    return flowContext;

1.1.0 - October 24 2025

🎨 Interactive Diff Component

A powerful new form component that allows users to review and selectively accept or reject text changes:

Interactive Mode: Click checkboxes and toggles to include/exclude specific changes
- Additions: Include or exclude new content
- Deletions: Apply or keep original text
- Replacements: Choose between previous or new text
Toolbar Controls:
- Zoom in/out for better readability
- Accept All / Reject All buttons for bulk operations
- Customizable button symbols and positioning
Real-time Updates: Result text recalculates as selections change

Example usage:

- step: FORM
  form:
    diffField:
      type: 'diff'
      mode: 'words'
      prevText: '{{ flowContext.original }}'
      text: '{{ flowContext.improved }}'
      interactive: true
      showDeletions: true
      diffControls:
        position: 'top'
        zoom: true
        acceptRejectAll: true

Perfect for AI text improvement workflows, translation review, and grammar correction tools.

⚡ Inline Steps Feature

Improved user experience for UI steps with background processing:

FORM, USER_SELECT, and SHOW_RESPONSE now support inlineSteps
Modal displays immediately with loading indicator
Background steps execute while modal is visible
Results seamlessly populate the UI when processing completes
Customizable loading messages via loadingText configuration
Error handling displays errors in-place within the modal

Example usage:

- step: USER_SELECT
  inlineSteps:
    - step: HUB_COMPLETION
      instruction: 'Generate 6 creative headlines for: {textContent}'
      response_format: 'list'
  promptText: 'Select a headline'
  enableKeyboardControl: true

This eliminates perceived wait time for long-running LLM calls and provides immediate user feedback.

🔧 Swing Integration System

Native integration with Swing Web Client UI components:

inlineMenu: Add buttons to the inline editor toolbar
- Appears when editing content in report, story, or DWP contexts and selecting text.
- Supports conditional visibility and enablement via expressions
- Configuration: label, icon, allowReadOnly, isEnabled, isActive
objectAction: Add actions to the object three dots menu (e.g. in search results or preview)
- Works across a lot of contexts
contextMenu: Add items to right-click context menus
- Appears in Explorer and other contexts
- Same configuration options as objectAction

Example configuration:

- title: 'Shorten Selection'
  swing:
    inlineMenu:
      enable: true
      icon: 'fa fa-compress'
      label: 'Shorten'
      allowReadOnly: false
      isEnabled: "{{ ctx.activeDocument.getSelection().getNode().elementName == 'p' }}"
  flow:
    - step: HUB_COMPLETION
      instruction: 'Shorten this text: {selectedText}'
    - step: REPLACE_TEXT
      at: CURSOR

⚠️ BREAKING CHANGE: New Deployment Structure

ProActions now uses a single unified bundle for deployment. The old multi-file structure is no longer supported:

Old Structure (v1.0.x):

~/methode-servlets/conf/swing/com.eidosmedia.swing.web-app/app/plugins/
├── pro-actions-core/
│   ├── pro-actions-bundle.js
│   ├── pro-actions-command.js
│   └── pro-actions-config.js
└── pro-actions-ext-ai-kit/
    └── proactions-ai-kit.js

New Structure (v1.1.0+):

~/methode-servlets/conf/swing/com.eidosmedia.swing.web-app/app/plugins/
└── pro-actions/
    ├── pro-actions-{version}.js    (single bundle file)
    └── pro-actions-config.js       (configuration file)

Migration Steps:

Backup configuration:
- Save a copy of plugins/pro-actions-core/pro-actions-config.js
Remove old files:
- Delete plugins/pro-actions-core/ directory
- Delete plugins/pro-actions-ext-ai-kit/ directory
Install new bundle:
- Create plugins/pro-actions/ directory
- Copy pro-actions-{version}.js from the release package
- Copy your backed-up pro-actions-config.js to plugins/pro-actions/
Verify configuration:
- Review pro-actions-config.js for any path references (usually none needed)
- Restart your Swing client to load the new bundle

Benefits:

Simplified deployment with single bundle file
Reduced complexity in installation process
Easier version management and updates
Consistent file structure across installations

⚠️ TEXT_MENU Deprecation

The TEXT_MENU component is now deprecated in favor of Swing's native inlineMenu integration:

Using TEXT_MENU in recent Swing versions will show both the inline toolbar and TEXT_MENU
The two components are not compatible and create a poor user experience
Action Required: Migrate your TEXT_MENU configurations to use swing.inlineMenu
TEXT_MENU will be removed in a future version

Migration example:

# OLD - Deprecated
TEXT_MENU:
  story:
    items:
      - type: 'button'
        icon: 'fa fa-icon'
        title: 'My Action'
        flowRef: 'My Action'
    # ...

# NEW - Recommended
title: 'My Action'
swing:
  inlineMenu:
    enable: true
    icon: 'fa fa-icon'
    label: 'My Action'

🤖 Enhanced Chat Completion Steps

Significant improvements to all chat completion steps (OPENAI_COMPLETION, HUB_COMPLETION, etc.):

Tool/Function Calling:

Define custom functions that the LLM can invoke during conversation
Functions can execute ProActions workflow steps, scripts, or templates
Support for required and optional parameters
Multi-turn conversations with tool results

- step: OPENAI_COMPLETION
  instruction: 'Ask the user a question using the tool askUser'
  functions:
    - name: askUser
      description: 'Ask the user a question'
      required_params: ['question']
      steps:
        - step: USER_PROMPT
          promptText: '{{ flowContext.question }}'

Binary Input/Output Support:
- Audio input support for speech-capable models
- Audio output support with automatic playback
- Image input improvements with better multimodal handling
- Base64 and blob handling for all media types
Structured Outputs:
- Enhanced response_format with json_object option.
Configuration Improvements:
- Support for conversation history via messages array

📚 Prompt Library Integration

Reuse and centrally manage prompts:

Reference stored prompts using promptId configuration
Prompts managed in Swing Prompt Management UI
Supports both system and user prompts
Variable resolution works within stored prompts
Reduces duplication and improves governance

Example:

- step: OPENAI_COMPLETION
  promptId: 'd112a306-8f7a-43bc-8f38-94161b6a91cd'
  # All other configurations remain available

🔄 New Control Flow Steps

Advanced workflow control capabilities:

BREAK: Exit loops conditionally

- step: WHILE
  condition: '{{ flowContext.counter < 10 }}'
  do:
    - step: SET
      counter: '{{ flowContext.counter + 1 }}'
    - step: BREAK
      condition: '{{ flowContext.counter == 5 }}'

FOR: Numeric and array iteration

# Numeric iteration
- step: FOR
  var: 'i'
  start: 1
  end: 5
  do:
    - step: SET
      item: '{{ flowContext.i }}'

# Array iteration
- step: FOR
  var: 'item'
  items: '{{ flowContext.myList }}'
  do:
    - step: SET
      current: '{{ flowContext.item }}'

TRY: Error handling with catch blocks

- step: TRY
  try:
    - step: SET
      value: '{{ riskyOperation }}'
  catch:
    - step: SET
      error: '{{ flowContext.error }}'
      fallbackValue: 'default'

PARALLEL: Execute multiple step sequences concurrently

- step: PARALLEL
  steps:
    - # Branch 1
      - step: HUB_COMPLETION
        instruction: 'Generate headline'
    - # Branch 2
      - step: HUB_COMPLETION
        instruction: 'Generate summary'

🎵 New Steps

PLAY_AUDIO: Display a customizable floating audio player
- Supports data URLs, HTTP URLs, and base64 audio
- Configurable theme (dark/light or custom colors)
- Position control (9 anchor positions)
- Progress bar, volume controls, and time display
- Autoplay support with browser compatibility fallbacks
- Draggable player with Shadow DOM isolation
```
- step: PLAY_AUDIO
  in: '{{ flowContext.audioDataUrl }}'
  autoplay: true
  theme: 'dark'
  position: 'bottom-right'
  showProgress: true
  showVolume: true
```
BASE64_TO_BLOB: Convert base64 strings to blobs
- Essential for handling binary data in workflows
- Supports custom MIME types
- Integrates with other steps requiring blob inputs

📝 FORM Enhancements

formConfig: New configuration object for FORM step behavior

Control modal sizing and appearance (width, height, dialogSize)
Fullscreen mode support
Typography customization (fonts, line heights)

- step: FORM
  title: 'Large Review Form'
  formConfig:
    dialogSize: 'xl'
    height: '80vh'
    diffFontSize: '14px'
  form:
    reviewField:
      type: 'diff'
      mode: 'words'
      prevText: '{{ flowContext.original }}'
      text: '{{ flowContext.improved }}'

See Forms and Inputs Guide for complete formConfig options.

Textarea Resize: Enhanced textarea component
- rows configuration for initial height
- resize option: none, vertical, horizontal, both
- Better control over form layout
```
- step: FORM
  form:
    notes:
      type: 'textarea'
      rows: 10
      resize: 'vertical'
```

🎯 USER_SELECT Keyboard Control

enableKeyboardControl: Navigate selection lists with keyboard
- Arrow keys for navigation
- Enter to select
- Escape to cancel
- Improves accessibility and power-user workflows
```
- step: USER_SELECT
  enableKeyboardControl: true
  promptText: 'Choose an option'
```

🔐 Admin Actions

admin flag: Mark actions as administrative
- Hidden from regular command palette
- Requires typing the full action name to access
- Useful for configuration reload, cache clearing, debugging tools
- Prevents accidental execution of sensitive operations
```
- title: 'Reload Configuration'
  admin: true
  flow:
    - step: # Reload steps
```

✨ Conditional Visibility with Templates

isAvailable now supports Nunjucks template expressions

Use {{ }} syntax for dynamic conditions

- title: 'Report Action'
  isAvailable: "{{ ctx.activeObject.getInfo().type == 'EOM::CompoundStory' }}"

Additional Improvements

JSON schema updates for all new features
New document lifecycle operations available in workflows and scripts:
- SAVE_DOCUMENT step supports optional close behavior
- CLOSE_DOCUMENT step closes the current document
- Scripting API now supports client.saveDocument({ close: true }) and client.closeDocument()

1.0.12 - February 7, 2025

🎯 Slash Menu

The new slash menu brings Notion-style command access to ProActions:

Type / anywhere in the editor to open the slash menu
Configuration consistent with TEXT_MENU and ONECLICK_MENU
Provides quick access to actions without keyboard shortcuts

New Steps

END_IF - Conditionally terminate flow execution
SANITIZE - Validate and repair XML content
- Essential for AI-generated XML where validity is uncertain
- Can automatically fix common XML issues
ELEVENLABS_TTS - ElevenLabs Text-to-Speech integration
ELEVENLABS_STT - ElevenLabs Speech-to-Text integration

Enhanced Features

HUB_YOUTUBE_UPLOAD
- Now supports uploading video thumbnails
- Improved video upload workflow
Méthode XML to Markdown
- New client.xmlToMarkdown(xml) function
- Simplifies content export and preview

Prime 8 Integration

Full integration with Prime 8 Command Palette framework
Specialized visualization for Prime 8
CHANGE_VIEW_SIZE step to control palette sizing

ProActions Financial Services

New client functions for financial services integration:

client.getContainerTextContent() - Fetch full report text in Swing and Prime
client.getContainerXmlContents() - Retrieve XML and LOIDs for all report sections
Enhanced client.getTextContent() - Returns selected section text
Enhanced client.getXmlContent() - Returns selected section XML

Improvements

client.isReadonly() now checks if active user holds the lock on the active object

1.0.11 - April 28, 2025

🎓 Schema Support

Created and registered ProActions AI-Kit schema for enhanced development:

Auto-completion for configuration files
Real-time validation
Inline documentation
Registered at schemastore.org
Reduces configuration errors significantly

Configuration Enhancements

Actions and Sections:

app attribute - Specify if action/section is for "swing" or "prime"
category attribute - Set sub-category for better organization in Prime Panel
hiddenInPanel attribute - Hide specific actions in Prime Panel
sectionIcon attribute - Custom icon for sections in Prime Panel
panelContext attribute - Specify context actions for selected elements

YouTube Integration

New steps for YouTube video management:

HUB_YOUTUBE_AUTH_INIT - Initialize OAuth 2.0 authentication
HUB_YOUTUBE_AUTH_STATUS - Check authentication status
HUB_YOUTUBE_AUTH_LOGOUT - Logout authenticated accounts
HUB_YOUTUBE_UPLOAD - Upload videos to YouTube

User Data Persistence

New client functions for user profile data:

getUserData() - Retrieve stored user data
setUserData() - Save user-specific data for future sessions

Development Support

Training Mode - Load configs from local development server
Enable with trainingMode configuration
Use with proactions-dev

Prime Enhancements

Improved XML management with getXmlAtXpath and replaceXmlAtXpath
Better XML insert location support
Improved read-only state detection using IsModifyable()

Step Improvements

CLIPBOARD - Browser compatibility fallbacks
FORM
- Key-value pairs for select options and radio items
- New form components: datetime, color, range
REST - formData attribute for multipart form data

Branding

Product name changed to "Eidosmedia ProActions" (formerly "Swing ProActions")

1.0.10 - February 27, 2025

📝 Enhanced Forms

New FormComponents for richer user interfaces:

headline - Section headers in forms
html - Rich HTML content display
markdown - Markdown rendering in forms
diff - Side-by-side text comparison
hr - Horizontal rules for visual separation
Custom buttons - Define multiple action buttons

New Steps

MARKDOWN_TO_HTML - Transform Markdown to HTML format

Enhanced Steps

USER_SELECT
- infoTitle attribute for additional context
- infoText attribute for descriptive information above list
SCRIPTING
- Now supports asynchronous JavaScript execution
- Greater workflow flexibility
SHOW_RESPONSE
- HTML response rendering support

Language Support

✨ Spanish Translation - ProActions now available in Spanish
Supported languages: English, Spanish, French, Italian, German

Context Handling

Multiple contexts support in sections and actions
Improved fallback mechanisms for Swing 5.2022.10
Context support added for Prime
Better context detection across platforms

Editor Protection

Steps that modify content now check for read-only mode:
- INSERT_TEXT, INSERT_XML
- REPLACE_TEXT, REPLACE_XML
- INSERT_LIST

Text and Oneclick Menu

forElement configuration now supports "DummyText"

Maintenance

Updated all dependencies to latest versions

1.0.9 - January 26, 2025

DeepL Write Integration

New Step: DEEPL_WRITE - Send text for improvement to DeepL
Enables text refinement and optimization
Note: DeepL Write API does not support XML format text

INSERT_XML and INSERT_LIST Enhancements

New position parameter for flexible insertion:

insertInline (default) - Insert at cursor position
insertBefore - Insert before cursor context
insertAfter - Insert after cursor context
Effective in Swing when at is set to "CURSOR"

1.0.8 - December 13, 2024

📊 Progress Bar Element

Visual feedback for long-running workflows:

SHOW_PROGRESS - Display progress bar
UPDATE_PROGRESS - Update progress state
HIDE_PROGRESS - Hide progress bar
Auto-hides on workflow completion or error

Dashboard Actions

AI-Kit now supports dashboard actions
New context configuration for actions and sections
Ensures actions only visible in relevant contexts
⚠️ Action Required: Add context: config for context-specific actions

Translation Management

Translations now built into solution
⚠️ Action Required: Remove standalone locale files

One-Click Menu Enhancement

New forTags configuration option
Define menu appearance based on parent tags

Content Filtering

Optional parameters for client.getTextContent and client.getXmlContent
GetTextContentOptions - Filter returned content
GetXmlContentOptions - Filter returned XML
Example: Exclude specific tags from content

Minor Updates

Hide sections - Hide all actions within a section
Azure OpenAI - Updated IMAGE_GENERATION API version to 2024-10-01-preview
client.getContext() - New function to identify user's current context

1.0.7 - November 14, 2024

🎨 Context-Based Menus

Two new highly configurable menus:

Text Selection Menu:

Appears when text is selected
Quick access to text-specific actions
Fully customizable per context
Permission-based visibility

One-Click Menu:

Appears on left side of content items
Activated when cursor is placed inside content
Context-sensitive actions
Per-action permissions

External Configuration Management

!include directive - Include external SysConfig files
Modular configuration management
Better organization and maintainability
Example: !include /SysConfig/ProActions/services.yaml

New Steps

CLEAR_SELECTION - Clear text selection without replacement
Useful before INSERT_TEXT or INSERT_XML steps

Stability AI Enhancement

SEARCH_AND_RECOLOR - Identify and recolor image components
Uses Stability AI's recoloring model

LLM Response Format

response_format option for OpenAI ChatGPT and Google Gemini
Enforce JSON response structure
Custom schema support
Alias list for JSON list enforcement

Step Improvements

DOWNLOAD - Now supports downloading blobs
INSERT_LIST / TO_LIST - Improved list parsing logic
- Removes double newlines
- Cleaner list generation

1.0.6 - October 11, 2024

🎤 Speech Services

New Text-to-Speech (TTS) models from OpenAI and Azure OpenAI
High-quality voice synthesis
Multiple voice options

Document Metadata

New placeholders for document name and issue date
Enhanced client functions for metadata access

🔒 Permissions System

Configure access control for sections and actions:

forUsers - User-level permissions
forGroups - Group-level permissions
forTeams - Team-level permissions

New Steps

DEBUG - Write flowContext to developer console
Creates breakpoint for debugging
Essential for workflow troubleshooting

Templating Engine

Integrated Mozilla's Nunjucks templating engine
Enhanced string handling
Powerful placeholder system

Step Enhancements

IF, WHILE, SWITCH - New condition attribute
- Supports string placeholders with Nunjucks
- More flexible conditional logic
SET, SWITCH - Simplified syntax
- Replaced name/expression attributes with key-value pairs
- Cleaner configuration

User Experience

Instant Actions - Execute on ProActions icon click
Improved workflow initiation

Technical Updates

Removed deprecated DOMSubtreeModified MutationEvent
Better cursor location monitoring

1.0.5 - September 3, 2024

New Steps

DOWNLOAD_TEXT - Download text content from flow
WRITE_CLIPBOARD - Write to user clipboard

FILE_UPLOAD Enhancement

New output types: text and arrayBuffer
More flexible file handling

Dynamic Placeholders

reportId - Report identifier
reportLoid - Report LOID
reportUuid - Report UUID

API Enhancements

Access Metadata, System Metadata, Usage Tickets, Virtual Attributes
Support for both story and report (container) contexts

1.0.4 - August 15, 2024

Performance Improvements

Faster placeholder resolution
Asynchronous configuration loading
Improved overall responsiveness

1.0.3 - August 14, 2024

🎨 Stability AI Integration

New steps for image processing:

STABILITY_AI_UPSCALE - Enhance image resolution
STABILITY_AI_OUTPAINT - Extend image boundaries
STABILITY_AI_SEARCH_AND_REPLACE - Find and replace in images

Placeholder Improvements

Option to omit text when variable cannot be resolved
More graceful handling of missing variables

Flow Control

Continue flow even when then or else block is missing
More resilient conditional execution

INSERT_LIST Enhancement

Add entries to existing lists
More flexible list management

REST Step Enhancement

Store response object in variable
requestOptions configuration
More control over REST calls

API Functions

getSiteConfigString() - Read from siteConfiguration
Better integration with Méthode configuration

New Placeholders

uuid - Document UUID
loid - Document LOID
appName - Application name (swing/prime)

Migration Notes

Upgrading to 1.1.0

⚠️ TEXT_MENU Deprecation (Action Required)

Migrate all TEXT_MENU configurations to use swing.inlineMenu
Using TEXT_MENU in recent Swing versions causes UI conflicts
TEXT_MENU will be removed in a future version

Migration steps:

Identify all actions using TEXT_MENU configuration
Replace with swing.inlineMenu configuration
Test in your Swing environment
Remove TEXT_MENU sections from configuration

New Features

All new features are opt-in and backward compatible
Interactive diff available in FORM step with type: 'diff'
Inline steps can be added to existing FORM, USER_SELECT, and SHOW_RESPONSE steps
Swing integrations work alongside existing action definitions
New control flow steps (BREAK, FOR, TRY, PARALLEL) available immediately

Recommended Actions

Consider using inlineSteps for actions with LLM calls to improve UX
Explore interactive diff for text improvement workflows
Review admin actions and add admin: true flag where appropriate
Update isAvailable conditions to use Nunjucks templates for better readability

Upgrading to 1.0.12

No breaking changes
New features are opt-in
Consider enabling Slash Menu for improved UX

Upgrading to 1.0.11

Schema support requires VS Code YAML extension update

Upgrading to 1.0.8

Remove standalone ProActions locale files (now built-in)
Add context: configuration to context-specific actions

Getting Help

Documentation: See Getting Started Guide
Issues: Check existing issues or report new ones
Community: Join ProActions Developer Community
Support: Contact Eidosmedia support team

Recent Highlights​

What's New in 1.2.0 (upcoming)​

What's New in 1.1.0 (latest)​

Major Features Since 1.0.0​

Version History​

Detailed Release Notes​

1.2.0 - March 2026​

1.1.0 - October 24 2025​

1.0.12 - February 7, 2025​

1.0.11 - April 28, 2025​

1.0.10 - February 27, 2025​

1.0.9 - January 26, 2025​

1.0.8 - December 13, 2024​

1.0.7 - November 14, 2024​

1.0.6 - October 11, 2024​

1.0.5 - September 3, 2024​

1.0.4 - August 15, 2024​

1.0.3 - August 14, 2024​

Migration Notes​

Upgrading to 1.1.0​

Upgrading to 1.0.12​

Upgrading to 1.0.11​

Upgrading to 1.0.8​

Getting Help​

Recent Highlights

What's New in 1.2.0 (upcoming)

What's New in 1.1.0 (latest)

Major Features Since 1.0.0

Version History

Detailed Release Notes

1.2.0 - March 2026

1.1.0 - October 24 2025

1.0.12 - February 7, 2025

1.0.11 - April 28, 2025

1.0.10 - February 27, 2025

1.0.9 - January 26, 2025

1.0.8 - December 13, 2024

1.0.7 - November 14, 2024

1.0.6 - October 11, 2024

1.0.5 - September 3, 2024

1.0.4 - August 15, 2024

1.0.3 - August 14, 2024

Migration Notes

Upgrading to 1.1.0

Upgrading to 1.0.12

Upgrading to 1.0.11

Upgrading to 1.0.8

Getting Help