Boston 4.6.0

📅 Release Date: July 18, 2025 - 📅 Next Release Date:

Release Summary

Autom Mate 4.6.0 “Boston” is a landmark release that fuses AI-native capabilities with operational robustness to redefine the automation experience.

This version introduces Autom Mate’s first generation of AI Agent features, enabling users to build, test, and deploy conversational agents with real-time chat, agent-specific settings, and Playground-based simulations. Bi-directional agent integration empowers workflows with dynamic data exchange and context-aware automation execution — bridging human intent and system action like never before.

On the platform side, real-time validation of actions, flow-wide error detection, and custom action version control give developers instant feedback and safer debugging tools. The addition of Microsoft Account Login streamlines enterprise access, while UI enhancements such as collapsible panels, improved variable visibility, and a centralized Errors screen elevate daily usability.

Critical system issues — including Base64 parsing failures, broken credential references, metadata residue in conditional logic, layout-breaking overflows, and inconsistent pagination — have been comprehensively resolved. These fixes improve system integrity, especially under enterprise-scale load.

Performance improvements span both backend logic and frontend rendering, with notable speed gains in large Autom lists and flow construction. From AI trials to action disablement, from alert management to help menu redesign — Autom Mate 4.6.0 offers a faster, smarter, and more secure platform for automation creators.

With this release, Autom Mate is not just stable and scalable — it’s intelligently extensible.

---

Highlights of this Release

🅱️ Bug Fixes

• Base64 Decode Not Handling Unpadded Inputs – The decoder now accepts Base64 strings without = padding and fails gracefully with informative errors when input is invalid.

• If Condition variableType Bug & Missing Variables – Residual metadata is now removed during type conversion, resolving hidden errors when switching from boolean to string conditions.

• Datatable Format Strips Static Values from Variables – Static content like [New] is now retained when combined with variables (e.g., ##xVariable##[New]).

• Deleted Credentials Still Functional in Automs – Automs now fail securely if a referenced credential has been deleted, preventing unauthorized or unintended access.

• OAuth Credential Name Lock Bug – Incomplete OAuth creation attempts no longer reserve credential names or leave behind invalid records.

• Overflow in Alert and Actioneer Profile Descriptions – Long, unbroken text now wraps properly, and a character limit has been introduced to prevent layout breakage.

• Pagination Input Triggers Premature Page Load – Page now navigates only after pressing Enter or clicking the Go button, not after each keystroke.

• System Action Fields Break \n Format on Update – Multiline text fields now preserve newline structure consistently during create and update operations.

• JSON Unmarshal Error in Email Send Action (Repeat Loop) – Email attachments now retain proper structure across all loop iterations, preventing type mismatch errors.

• MateDrive Attachment Parsing Fix – Email Send action now correctly handles mate:/ paths in the attachments field without throwing string parsing errors.

---

🛠️ Enhancement

• Autom Screen Performance Optimization – 20% faster load and interaction speed in large-scale automation flows.

• Quick Access in Autom List – Navigate frequently used Automs faster from the list view.

• Improved Help Menu – Access more in-app resources including FAQs, how-tos, and documentation.

• Error Configuration Tab – Simplified management of how and when errors are captured and reported.

• Action Disable Option – Temporarily disable specific actions or groups to support safer testing.

• Alert Management Simplification – Enhanced alert configuration tools for easier monitoring setup.

• Disable Non-Dev Mode Features – Prevents any interaction or modification when Dev Mode is off, enabling a secure, read-only preview experience.

• Data table Library Model View Changes – Introduces a reusable and responsive DataTable UI with ellipsis-based overflow handling, tooltips for hidden data, and full-screen mode support.

• Disable SSL Configuration UI and Redirect Users to Support – The SSL section in settings is now disabled and guides users to submit a support request via Xurrent.

• Dev Mode Switch: Toggle Dotted Background in Autom Builder – Canvas background now adapts between clean and dotted views based on Dev Mode state.

• Remove "Delete" Option from Autom Context Menu – Streamlines individual item menus and reduces accidental deletion risk.

• License-Aware Import Validation – Imports now trigger validation checks for license-restricted libraries, preventing unauthorized action usage.

• Removal of Obsolete Language Flag – The unused language attribute in Autom metadata has been removed to simplify the system structure and improve data integrity.

• Default Panel Collapsing on Autom Load – The Autom Flow screen now loads with both side panels (Library and Configuration Form) collapsed, offering a cleaner and distraction-free workspace.

• Improved Variable Icon Visibility – Variable icons are now visible in the library and Data Manager for better navigation and usability.

• Variable Tab in Action Panel – Adds a dedicated tab for managing variables separately from actions, improving clarity and workflow organization.

• Live Error Indicator on Autom Screen – Displays the current number of validation errors and offers direct navigation to the Errors panel for faster troubleshooting.

• Errors Screen UI Design – A centralized validation screen that displays categorized flow issues, supports direct navigation, and updates in real time.

• Monitoring Logs: Action Redirect Button – A redirect button has been added to each log entry in Monitoring Logs, allowing users to jump directly to the related action within the Autom Builder. This enhancement significantly improves debugging efficiency by reducing time spent searching for the source of issues.

• Custom Action Version Sync – Updating a custom action version now reflects immediately in both the flow UI and JSON definition, ensuring runtime consistency.

• Custom Action Version Switch UI – Enables developers to switch between available custom action versions directly from the Autom Builder interface with real-time feedback.

---

✨ New Features

• Bi-Directional AI Agent Integration – Enables two-way data exchange between AI agents and Autom Mate for dynamic workflow automation.

• Agent Details Panel – View and configure agent credentials, settings, and behaviors from a central UI.

• AI Agent Builder – Create, edit, and manage AI agents without affecting core Autom flows.

• Playground for Testing AI Agents – Simulate conversations and test agent behaviors in a safe, isolated environment.

• Real-Time Chat Integration – Interact with AI agents using live chat interfaces embedded in the platform.

• Chat History Storage – Preserve agent interaction history for context-aware automation.

• AI Agent List Interface – Users can now view and manage AI Agents from a centralized, responsive listing page, featuring real-time data, pagination, and secure access.

• Deploy AI Agent from Details Page – Users can now deploy their AI Agent configurations directly from the Agent Details screen with a single click.

• Autom Creation from AI Agents – AI Agents can now dynamically create new automation flows inside Autom Mate, accelerating agent-based task orchestration.

• AI Agent 30-Day Trial Enablement – Allows users to activate and monitor a time-limited trial of AI Agent functionality, with clear expiration visibility and sales engagement CTA.

• Real-Time Action Validation – Every newly added action is now validated instantly. Any errors (missing required fields, incorrect values) are listed in the Errors panel and visually marked in the flow editor.

• Custom Action Version Listing Service – Backend service to fetch all downloaded versions of a custom action, enabling version switching in developer mode.

• Initial Flow Validation – All actions in a flow are automatically validated when the flow is opened, helping identify issues early.

• Microsoft Account Login – Users can now securely log in with their Microsoft work, school, or personal accounts.

• Microsoft Sign-In Button on Login Screen – A new “Sign in with Microsoft” button has been added to the login screen, allowing users to log in with their Microsoft accounts through OAuth 2.0. This improves accessibility and streamlines login for enterprise users.

---

Detailed Version

🅱️ Bug Solution

• Base64 Decode Not Handling Unpadded Inputs Previously, Base64 decoding in Autom Mate required manual padding (=) to function correctly. Valid but unpadded Base64 strings—such as NG1lLnFhL0NpLzk4MDMxMzM—would fail silently or return non-informative errors. With this fix, the decoder now automatically handles padding internally and ensures accurate decoding for all valid Base64 formats. Additionally, when truly invalid input is provided, the system returns a clear, context-aware error message to aid debugging. Unit tests have been added to verify padded and unpadded input handling, and no regressions were observed in dependent modules.

• If Condition variableType Bug & Missing Variables A bug was causing conditions to retain outdated metadata when the field type was switched from boolean to string—specifically, the variableType: "boolean" key would remain in the condition object even after type change. This led to unexpected runtime behavior and missing variable mappings. The issue typically occurred when a condition was created with a boolean variable, then updated to string type using the same field (e.g., source, target). The system now correctly removes or updates the variableType property based on the current type. Automated tests were added to validate this logic across condition creation, updates, and variable reuse. Existing flows are unaffected, and improved error messaging has been implemented to assist users in identifying type-related misconfigurations.

• Datatable Format Strips Static Values from Variables A bug was identified where combining a variable with static text in a Datatable field—such as ##xVariable##[New]—resulted in only the variable’s value being displayed. The static portion ([New]) was being stripped during rendering, leading to loss of formatting context and user-defined annotations. This release corrects the parsing logic so that both the variable and the adjacent static text are rendered together, as intended. The fix has been verified using a provided Autom export and tested with various combinations of variable and text input. Unit tests have been added to prevent regression in future builds.

• Deleted Credentials Still Functional in Automs Previously, Autom Mate did not properly validate the existence of credentials at runtime. If a credential was deleted after being configured in an Autom, the automation could still execute using the credential’s cached data. This posed a potential security and integrity risk. With this fix, credentials are now fully unlinked and invalidated at the time of deletion. Autom executions will perform a runtime check to confirm that any referenced credential still exists and is valid. If not, the system will return an informative error, halting execution. This behavior is now enforced across all credential types—OAuth, Basic Auth, API Key, etc. Unit tests and integration checks were added to ensure deleted credentials are no longer accessible under any condition.

• OAuth Credential Name Lock Bug Previously, if the OAuth credential creation process failed or was canceled mid-flow, the system retained a partial credential record—including its name—despite not storing a valid token or completing configuration. This prevented users from creating a new credential with the same name, leading to confusion and blocking valid retry attempts. With this release, incomplete or failed OAuth credential entries are automatically cleaned up. The system now ensures that credential names are only reserved upon successful completion of the creation process. Edge cases—such as disconnected tokens, empty records, or browser cancellations—are now handled gracefully. Logging has been added for failure and cleanup events, and new unit/integration tests ensure the robustness of both successful and unsuccessful flows.

• Overflow in Alert and Actioneer Profile Descriptions Previously, when users entered long text strings without spaces in the Description fields of the Alert Profile and Actioneer Profile pages, the content would overflow its container and disrupt the layout. With this fix, the UI now enforces automatic word breaking for unspaced text that exceeds the maximum width. Additionally, a character limit has been implemented on the Description input fields to prevent excessive input lengths. These changes ensure consistent alignment, prevent visual breakage, and enhance overall user experience in profile management interfaces.

• Pagination Input Triggers Premature Page Load Previously, when users typed into the page number input within pagination controls, the interface would reload the page after each digit entered—causing abrupt jumps and making navigation nearly unusable for large page numbers. This update resolves the issue by deferring navigation until the user either presses Enter or clicks a new “Go” button placed next to the input field. The input now accepts multi-digit entries as expected, and only final confirmation triggers the page change. This fix ensures smoother and more predictable pagination behavior across all relevant views.

• System Action Fields Break \n Format on Update A bug was identified where the note field in system actions handled newline characters (\n) inconsistently. While line breaks were correctly preserved during create requests, they were being altered or lost during update requests, resulting in broken formatting and inconsistent display. With this fix, both backend processing and UI representation have been aligned to treat newline characters consistently. Multiline content now maintains its structure during both creation and update flows. Automated test coverage has been expanded to verify this behavior at both API and UI levels.

• JSON Unmarshal Error in Email Send Action (Repeat Loop) A bug was identified where the Email Send action inside a repeat loop would work correctly on the first iteration but fail on subsequent iterations with the error: json: cannot unmarshal string into Go value of type model.Attachment.

  The root cause was that the attachment input, initially structured as a valid JSON object or array, degraded into a plain string in later iterations—likely due to improper data handling in upstream steps such as REST responses or Regex outputs. This caused a type mismatch when the system attempted to deserialize the input into the expected model.Attachment structure.

  With this fix:

  • The attachment input is now consistently validated and parsed across all iterations.

  • Defensive type-checking and JSON normalization have been added before Email Send execution.

  • Flow stability is restored when sending emails with attachments inside loops.

  The change is backward-compatible and includes extended test coverage for loop-based scenarios involving file or data attachments.

• MateDrive Attachment Parsing Fix Resolved an issue where entering file paths copied from MateDrive (e.g., mate:/asd/file.json) into the Email Send action's attachments field would result in a parsing error: invalid character 'P' in string escape code. The system now properly recognizes and processes MateDrive file paths as attachments. If an invalid or unresolved path is provided, the system returns a clear and user-friendly error message such as "Invalid MateDrive path." This ensures smoother email delivery using MateDrive-based files.

🛠️ Enhancement

• Autom Screen Performance Optimization Performance tuning has reduced latency and improved rendering speeds, especially for large automation projects. This leads to a smoother user experience during design and execution.

• Quick Access in Autom List Users can now quickly locate and open frequently accessed Automations directly from the Autom list without additional clicks or search.

• Improved Help Menu The Help menu now includes enhanced documentation, onboarding guides, and troubleshooting content directly within the interface.

• Error Configuration Tab The new tab within Agent Details allows users to define error logging and response strategies, including retry policies and failure alerts.

• Action Disable Option Actions and action groups can be selectively disabled during design and testing. This gives developers more control over iterative changes.

• Alert Management Simplification The alert setup process has been restructured to be more intuitive, allowing quicker configuration of rules, thresholds, and delivery methods.

• Disable Non-Dev Mode Features With this release, Autom Mate introduces a robust preview-only mode that is automatically activated when Dev Mode is turned off. In this state, users can no longer modify automations—actions like dragging, adding, deleting, renaming, or configuring elements are disabled. Access to tools such as MateDrive, Debug Mode, BI Estimator, and variable or credential editors is restricted, while UI components like the Publish button are visually dimmed to reflect their inactive status. The goal is to ensure that the platform remains in a safe, read-only mode, allowing users to preview flows without making any unintended changes.

• Data Table Library Model View Changes This release introduces a shared Data Table component used within action configuration forms to present structured tabular data more effectively. A defined content limit ensures that when the visible area is exceeded, data is truncated with an ellipsis, and a tooltip appears indicating the number of additional hidden rows or columns. The component includes a full-screen view option for improved readability. These enhancements have been implemented in actions such as “Data Table – Insert Row” and “Assign.” Existing behavior in actions like “Data Table – Merge Columns” remains unchanged. QA has confirmed that data alignment remains top-bound in full-screen mode, input fields maintain their sizing during edits, and no overflow occurs in the Action Control Center.

• Disable SSL Configuration UI and Redirect Users to Support In this release, the SSL certificate management section under Settings > Port & SSL has been disabled to prevent user interaction, as the functionality is not currently supported via the UI. The section is now grayed out and includes a clear message: "SSL management is currently unavailable via this panel. Please contact Autom Mate Support for SSL-related operations." A “Contact Support” button is also provided, which redirects users to the Xurrent support portal and, where possible, pre-fills the subject with “SSL Management Request.” The Import Certificate button is hidden or disabled accordingly. This update ensures users follow the correct path for SSL-related changes, while the rest of the Port configuration settings (Host Name, Web Port, Trigger Port) remain fully functional.

• Dev Mode Switch: Toggle Dotted Background in Autom Builder In this release, the Autom Builder canvas background becomes context-aware based on the Dev Mode toggle. When Dev Mode is enabled, the traditional dotted grid background appears, aiding visual alignment and structural planning. When Dev Mode is disabled (which is now the default state), the background switches to a clean, dot-free view to emphasize read-only or preview mode. The transition is seamless and instant—no reload is required—and is scoped strictly to the canvas area, ensuring that other UI components remain unaffected.

• Remove "Delete" Option from Autom Context Menu To reduce the likelihood of accidental deletions and simplify the user interface, the "Delete" option has been removed from the three-dot context menu shown next to each Autom item. Users can still perform deletions using the existing bulk delete option available in the top action panel when selecting multiple items. Other context menu actions—such as Edit, Duplicate, and Log History—remain fully available and unaffected. The UI layout has been preserved without any visual disruption following the removal of the delete item.

• License-Aware Import Validation To prevent the use of actions that are not supported by a user's license or system type, this release adds a validation check during the Autom import process. If a flow contains actions from a library not covered by the license, the import will fail with a clear message: “Import failed. ‘xxx’ action is not compatible with your current license or system type!” Additionally, any such unsupported actions—if previously imported—can no longer be copied or pasted into other automations. This change helps maintain license integrity and avoids user-side execution failures caused by unauthorized feature usage.

• Removal of Obsolete Language Flag The legacy language flag in Autom JSON definitions has been removed, as it no longer serves any runtime or UI purpose. This cleanup helps reduce unnecessary metadata, avoids confusion during Autom export/import, and ensures that Autom definitions remain aligned with current platform behavior. There is no functional impact on users or automation outcomes, but it improves maintainability and forward compatibility across environments.

• Default Panel Collapsing on Autom Load To improve initial focus and visual clarity, the Autom Flow screen now hides both the left-hand Library panel and the right-hand Configuration Form by default when loading an Autom. This enhancement reduces visual noise on first entry, particularly in complex flows, and allows users to engage with the canvas without distractions. Panels can still be expanded manually as needed, preserving full functionality while improving the default user experience.

• Improved Variable Icon Visibility Variable icons have been added to both the left-side library panel and the Data Manager interface. This enhancement helps users identify variable-related actions more quickly, improving visual clarity and usability during automation development. It also contributes to a more consistent UI layout across Autom Mate's workspace.

• Variable Tab in Action Panel Previously, variables and actions were grouped under the same Action Panel, which could cause visual clutter and hinder navigation. In this release, a dedicated Variables tab is introduced alongside the Actions tab within the Action Panel.

  • Users can now switch between Variables and Actions using tab navigation.

  • This update ensures a cleaner UI and better separation of concerns, especially in large or variable-heavy automations.

  • The change enhances discoverability and manageability of variables without interfering with the core action-building process.

  This improvement is particularly beneficial for users working on complex flows that rely heavily on variable manipulation.

• Live Error Indicator on Autom Screen This enhancement introduces a persistent error indicator icon on the Autom Builder screen, enabling users to quickly identify issues as they construct or edit flows.

  • The icon dynamically shows the number of validation errors present in the flow.

  • The error count updates in real time as users fix or introduce issues.

  • Clicking the icon opens the Errors screen or sidebar, allowing users to jump directly to problematic actions or configuration points.

  • A tooltip appears on hover, providing context and encouraging proactive issue resolution.

  • The feature is responsive and works seamlessly across different devices and screen sizes.

  This improvement significantly reduces the time required for debugging and enhances transparency in automation flow validation.

• Errors Screen UI Design This enhancement introduces a fully redesigned Errors screen in the Autom flow editor. It provides a centralized view for all validation issues in a flow and allows users to easily identify and resolve errors.

  • Displays categorized error types (e.g., Missing Fields, Invalid Configurations, Version Conflicts).

  • Each error links directly to the related action and highlights it in the flow.

  • Error list updates in real time as users fix or introduce issues.

  • Filtering and search allow for fast error lookup in complex flows.

  • Fully responsive and visually consistent with the Autom Mate design system.

  This improvement simplifies debugging workflows and supports better flow integrity management.

• Monitoring Logs: Action Redirect Button This enhancement introduces a highly practical feature to the Monitoring Logs interface by adding a redirect button next to each action-related log entry. When clicked, the button prompts the user with a confirmation message and, upon acceptance, navigates directly to the corresponding action in the related Autom flow.

  • ✅ Case 1: If the action exists, the user sees a confirmation prompt: “Are you sure you want to be redirected to the Autom Builder screen?”

  • ⚠️ Case 2: If the action no longer exists: “The action you are trying to view could not be found within the Autom. Would you still like to be redirected to the Autom Builder screen?”

  • 🔁 Case 3: For sub-actions within parent actions (e.g., Xurrent actions), the system opens the Autom Builder in a new tab with context-aware messaging: “This is a sub-action of 'Xurrent - View Request'. To view it properly, we’ll open the Autom Builder in a new tab and show you its parent action there. Would you like to continue?”

  This feature works reliably across different Autom states (Draft, Active), supports visual highlighting, includes hover tooltips, and provides fallback messaging when redirection cannot occur. All functionality has been validated through UI testing and is compatible across supported browsers.

• Custom Action Version Sync When a custom action’s version is updated from the UI, the selected version is now immediately applied to both the visual flow and the JSON definition. This ensures that execution always uses the intended version. Proper validation, error handling, and undo support are included.

• Custom Action Version Switch UI When Developer Mode is enabled, each custom action now includes a version-switch control. Clicking it displays a list of all available versions, allowing users to update to a different version seamlessly. The UI reflects the change instantly, with appropriate success and error feedback states. This enhancement improves control over version management and supports smoother iteration across action updates.

✨ New Features

• Bi-Directional AI Agent Integration Autom Mate now supports secure two-way communication with AI Agents through JWT token-based authentication. This allows for real-time data exchange and interaction, creating dynamic and adaptive automation workflows.

• Agent Details Panel Users can now manage each AI Agent through a dedicated interface. The panel allows configuration of agent-specific credentials, metadata, and runtime options to streamline setup and debugging.

• AI Agent Builder A no-code builder interface enables users to define agent behavior, channels, and integration logic without altering the main Hyperflow logic. CRUD operations are fully supported.

• Playground for Testing AI Agents This isolated test environment lets users simulate conversations and verify agent responses before production deployment. The playground supports iterative improvement of AI logic.

• Real-Time Chat Integration The chat screen now supports live interactions with AI Agents. Users can test, monitor, and operate flows through natural language interfaces.

• Chat History Storage All agent-user conversations are saved to DB, allowing users to view past interactions and retain session context.

• AI Agent List Interface This release introduces a dedicated AI Agent List page, providing users with a centralized and secure interface to browse, access, and manage AI Agents. Upon login, authenticated users are presented with a clean and responsive UI that displays each agent’s name, model, temperature setting, and last updated date. The page is optimized for usability, with support for loading states, empty list handling, and backend-driven data fetching through secure API calls. For environments with a large number of agents, the page includes pagination or infinite scroll, ensuring smooth performance. This foundational component enhances discoverability and manageability of agents across the platform.

• Deploy AI Agent from Details Page The Agent Details interface now includes a fully functional Deploy button, allowing users to persist the current configuration of their AI Agent to the Autom Mate platform. This feature supports both the creation of new agents and updates to existing ones, depending on whether an autom_id is present.

  When clicked, the button sends a secure, validated payload (including model, credentials, temperature, and name) to Autom Mate’s backend API. The deployment process ensures:

  • Proper authentication (via service key or JWT)

  • Error feedback if submission fails

  • Reliable agent persistence and immediate availability across the system

  This feature bridges the configuration UI with Autom Mate’s automation engine and makes it possible to operationalize agents in real-time.

• Autom Creation from AI Agents With this release, Autom Mate introduces backend support for AI Agents to create new automation flows autonomously. Through a secure API endpoint, structured payloads from AI Agent Builder or Playground are validated and translated into Autom Mate-compatible automation flows.

  This feature enables:

  • Seamless orchestration between LLM reasoning and Autom Mate actions

  • One-click Autom generation from chat interactions or predefined logic

  • Agent-authored flows that include dynamic trigger types, action sequences, and metadata mapping

  Security, validation, and auditing have been fully integrated to ensure:

  • Only authenticated agents can initiate Autom creation

  • Input formats and flow logic are validated before persistence

  • All actions are traceable with associated autom_id and autom_version_id references

  🔒 This capability supports future use cases like agent-driven task setup, scalable AI workflows, and fully autonomous automation initialization.

• AI Agent 30-Day Trial Enablement This feature introduces a seamless 30-day trial activation flow for AI Agent capabilities when no valid add-on license is present.

  • Upon first access to the AI Agent module without a license, users are presented with a “Start 30-Day Trial” button alongside a “Contact Sales” CTA.

  • Once the trial is activated, the system displays the number of remaining trial days directly on the page.

  • After the trial period ends, all AI Agent content becomes blurred to prevent interaction, and only the Contact Sales button remains accessible to facilitate conversion.

  This flow ensures license compliance while offering users the opportunity to explore advanced features in a controlled and transparent manner.

• Real-Time Action Validation

  This feature automatically validates each newly added action within an Autom flow. As soon as the action is inserted, the system checks for required fields, configuration mistakes, and unsupported values. If any issues are found:

  • They are listed in the Errors screen, with references to the corresponding action.

  • The problematic action is visually marked in the flow editor.

  • Validation updates in real time as the user edits the action.

  The system handles all action types consistently and ensures that no performance issues occur — even when adding multiple actions rapidly. This empowers users to catch errors early and build more stable flows with confidence.

• Custom Action Version Listing Service A backend service that returns all downloaded versions of a given custom action. It supports input validation, proper error handling, and is optimized for performance. This service enables version switching within the developer mode interface and is documented for frontend integration.

• Initial Flow Validation on Load When an Autom flow is opened, the system now validates all existing actions for configuration issues, missing required fields, and invalid settings. Errors are displayed in the Errors screen and flagged directly on problematic actions within the editor. This helps developers detect and resolve issues before any manual edits or execution. The validation process is optimized for performance and re-runs whenever changes are made to the flow.

• Microsoft Account Login Support Autom Mate now supports logging in with Microsoft accounts using OAuth 2.0. Both organizational (work/school) and personal accounts are supported. Key capabilities include:

  • Secure redirection to Microsoft login and callback handling

  • Decoding and validating id_token on the backend

  • Matching users by email to existing Autom Mate accounts

  • Whitelist enforcement to allow only authorized users

  • Session/token creation after successful login

  • Environment-based configuration for redirect URIs and scopes

  Edge cases such as missing claims, invalid tokens, or email mismatches are handled gracefully with proper error feedback.

• Microsoft Sign-In Button on Login Screen The Autom Mate login screen now includes a "Sign in with Microsoft" button, allowing users to authenticate using their Microsoft work, school, or personal accounts. Clicking the button initiates the OAuth 2.0 login flow via the configured Azure App. Upon successful authentication, users are logged into Autom Mate with appropriate permissions. The button is styled according to Microsoft’s branding guidelines and includes proper error handling for failed or canceled attempts. Existing login methods remain unaffected, and the updated flow is fully tested across browsers and screen sizes.

---

User Notices & Guidance

As part of the Boston 4.6.0 release, several behavior changes, improvements, and stricter validations have been introduced. To ensure a smooth transition and prevent confusion, please review the following usage hints, system warnings, and best practices:

Warnings (Critical Behavior Changes)

• Dev Mode Restrictions Apply When Dev Mode is turned off, users will not be able to drag actions, modify triggers, update variables, or rename Automs. This mode is read-only. To make any changes, Dev Mode must be enabled.

• Deleted Credentials Are Now Strictly Invalidated Credentials that have been deleted will no longer function in previously configured Automs. Execution will fail with an appropriate error. Review older flows to ensure they reference valid, active credentials.

• OAuth Credential Name Lock Bug Fixed Previously, failed or canceled OAuth credential attempts could reserve the credential name. This behavior has been fixed. Users who encountered this bug can now reuse those credential names.

• Email Send Action Inside Repeat Loop Requires Structured Attachment The Attachment field in the Email Send action must remain a valid JSON object (model.Attachment) throughout all iterations. If you use REST responses or regex outputs, ensure the format does not degrade into a string on later iterations.

• License-Based Action Restriction Now Enforced on Import Automs containing actions not permitted by the current license will now fail to import. When sharing .autom files across teams or customers, ensure the actions used are license-compliant.

• Live Error Indicator Now Affects Flow Readiness The new real-time error indicator on the Autom screen reflects validation issues in real time. If errors are present, flow execution may be blocked or produce unexpected results. Always resolve listed issues before attempting to run or publish an Autom.

• Initial Flow Validation May Surface Legacy Errors Flows created before version 4.6.0 may now show validation errors during initialization that were previously ignored. Review and update these flows accordingly.

• Action Version Switching Can Affect Runtime Behavior Switching the version of a custom action will immediately affect how that action behaves. If using older versions with different field sets or logic, test carefully before deploying.

• Microsoft Login Requires Email Match to Existing User Users logging in with Microsoft accounts must have a corresponding Autom Mate user record with the same email. If no match is found, login will fail even if Microsoft authentication succeeds.

• Microsoft Login May Appear Stuck After Success In some cases, after a successful Microsoft login, users may see a black screen with no redirection despite receiving a “success” message. This typically indicates a temporary redirect or token processing issue. If this occurs, refresh the page manually or contact your administrator. A fix is in progress.

• Modifying Agent-Generated Autom's

  When an AI Agent is created, Autom Mate automatically generates a linked Autom flow to handle the agent’s execution. Users are free to open, extend, or modify this flow — including adding new actions or adjusting logic. However, Changing the Trigger Method or key flow structure may cause the AI Agent to stop functioning as expected. We recommend carefully reviewing and testing the agent after any modification to its linked Autom.

---

Hints (Productivity Tips)

• Use the Agent Playground Before Going Live AI Agents can be safely tested in the Playground before deploying them to production environments. This sandbox allows simulation without impacting core workflows.

• Chat History Is Now Persisted Interactions with AI Agents are saved, providing context for follow-up actions or long-running workflows. Users can view and reuse chat histories as needed.

• Disable Actions for Partial Testing You can now disable individual or grouped actions to test parts of a flow without deleting them. This supports safer, modular testing and faster debugging.

• Explore the Enhanced Help Menu The in-app help menu has been expanded with quick access to FAQs, platform documentation, and how-to resources. Use this area to onboard faster or resolve issues independently.

• Pagination Input Now Requires Confirmation Entering a page number will no longer immediately trigger a reload. Users must press Enter or click the Go button to confirm navigation—reducing unintended jumps.

• Use “Variables” Tab to Organize Data The new Variables tab in the Action panel provides a centralized view of all variables in a flow. Use this to rename, inspect, and trace usage of variables more efficiently.

• Collapse Side Panels for Focused Building By default, the Library and Configuration side panels are now collapsed when opening a flow. Use this cleaner layout to stay focused, especially on large flows.

• Custom Action Version Sync Ensures JSON Consistency When a custom action is updated, the flow UI and underlying JSON are now automatically synchronized. This eliminates mismatches between design and execution.

• Monitoring Logs Offer Deep Linking You can now jump directly from a log entry to the related action within a flow using the new redirect icon — great for tracing issues faster.

---

Best Practices (Recommended Usage)

• Validate Base64 Inputs When Decoding The decoder now tolerates missing padding characters (=), but malformed inputs will raise structured errors. Always ensure your Base64 strings are complete and well-formed.

• Review Conditions When Changing from Boolean to String Changing a condition type from boolean to string will now automatically remove obsolete metadata. Still, it’s good practice to inspect the JSON for complex conditions after changes.

• Preserve Formatting in Datatables When combining variables with static text (e.g., ##xVariable##[New]), ensure syntax is respected. The system now retains both parts properly, but avoid nested or ambiguous delimiters.

• Avoid Long Unbroken Text in Profile Descriptions Description fields in Alert and Actioneer profiles now have a character limit and auto-wrapping enabled. For better readability, avoid using excessively long or space-free input.

• Review Variable Icons for Better Navigation Variable icons are now visible throughout the platform. Use them to identify dynamic fields quickly, especially when working with large data tables or conditional logic.

• Use AI Agent Trial Period Strategically The 30-day trial for AI Agents allows full usage during the test window. Evaluate high-value use cases (e.g., ticket creation, chatbot flows) before activating full access.

• Avoid Editing Automs in Read-Only Mode If Dev Mode is disabled, flow edits will not be saved and may lead to confusion. Always check Dev Mode state before making changes.

• Preserve Newline Formatting in Multiline Fields The system now preserves \n formatting consistently, but avoid injecting raw special characters (like \r, \t) into system action fields.