Product Newsletter — February 2026
The February 2026 edition covers updates across modules — including Quick Replies in AI Colleagues Studio that let users tap through common actions instead of typing, new Skill Invocation Controls for fine-tuning how skills execute within conversations, a redesigned Settings navigation in the Helpdesk for faster configuration, a new toggle to lock ticket ratings for data integrity, and improved failure reporting in the Workflows Studio that makes failed API/AIC submissions easier to find and act on.
In This Edition
| Module | Highlights |
|---|---|
| AI Colleagues Studio | Quick Replies, Skill Invocation Controls, UX improvements & bug fixes |
| Helpdesk | Redesigned Settings navigation, Disable Edit Rating control |
| Skills (Workflows) Studio | Simplified reporting for failed submissions |
AI Colleagues Studio
Updates to the AI Colleagues platform this month focused on making conversations faster for end-users and giving admins more control over how skills behave.
Quick Replies
The assistant can now present one-tap Quick Reply buttons for common actions — Yes/No confirmations, category selections, Approve/Reject decisions, and resuming in-progress forms. Instead of typing out responses for straightforward choices, users simply tap a button, making flows more predictable and reducing unnecessary back-and-forth. Quick Replies work across all supported channels including Web, Desktop, Mobile, MS Teams, Slack, and WhatsApp, with each channel rendering buttons in its native format (e.g., Adaptive Cards on Teams, Block Kit on Slack). Behind the scenes, clicking a Quick Reply triggers a restricted tool mechanism that bypasses the AI's standard decision-making to execute the intended skill immediately — ensuring fast, deterministic outcomes for routine interactions.
📖 Documentation: Quick Replies
Skill Invocation Controls
A new execution control setting is now available for skills. Admins can configure whether a skill must run fresh on every invocation or let the system optimise by reusing results from earlier in the conversation. This is particularly useful for skills that fetch data (like leave balances or approval statuses) where the result may or may not change within a single conversation session. The setting is available at the individual skill level, giving admins granular control over the freshness-vs-performance tradeoff for each capability.
📖 Documentation: Execution Control
UX Improvements & Bug Fixes
Several targeted fixes to improve the day-to-day admin experience in the AI Colleagues Studio:
- AOP Instructions Editor — Fixed overlay issues with the AOP/Skill selector and resolved a bug where the spacebar was breaking skill selection.
- AOP Version History — Fixed the sticky version history panel that was obstructing the editing view.
- MCP Server Skill Creation — Added a navigation CTA to Integrations > Custom MCP Server during MCP-based skill creation, making it easier to set up external integrations inline.
- Debug Console — Fixed the Debugging Console CTA for system-executed runs under Run History, so admins can now inspect automated executions properly.
Helpdesk
This month's helpdesk updates bring a cleaner configuration experience and a new control to protect the integrity of ticket feedback data.
Redesigned Settings Navigation
The Settings icon has moved to the bottom-left of the navbar. Clicking it now opens a persistent sidebar list instead of the previous tile view on the main page. This change aligns the Helpdesk with other product modules for a unified experience, reduces visual clutter in the configuration area, and lets admins jump between entities like Users, SLA, and Automation instantly without navigating away.
📖 Documentation: Admin Settings
Disable Edit Rating
A new toggle under Helpdesk > Settings > General allows admins to restrict users from re-rating a closed ticket. Once a ticket is rated during the closing process, that rating becomes locked — the only way to modify it is to reopen the ticket and close it again with a new selection. This ensures the integrity of feedback data by preventing accidental or unauthorized rating changes after resolution.
📖 Documentation: Configure Ticket Rating
Skills (Workflows) Studio
Simplified Reporting for Failed Submissions
Workflow submissions initiated via API or AI Colleagues that fail will now be classified under Failed instead of Pending Submissions — making them significantly easier to find and act on. These cases will also surface across all applicable reports, giving admins better visibility into submission health without needing to manually check pending queues. Failed submissions initiated through forms (not API/AIC) continue to appear under Failed Submissions as before.
