End users can now see information about tool and knowledge base call executions in the chat in a new look.

To prevent possible issues where messages would be passed to the model in the wrong order, extra logic was added so that an empty user message is added to a request (if none was there before).

We fixed a padding and a small bug where error states of tool calls were stored incorrectly.

A bug was fixed, where suggested user prompts were not inserted into the chat on click.

User Access and Approval is a new GenAI Commons and ConversationalUI feature to make tool calls visible to users and provide control for users to decide if tools should be executed or rejected. They can be either configured at runtime in AgentCommons or at design time using GenAI Commons actions when adding tools/functions to a request. For knowledge bases it can be decided whether they are visible to the user or not. Per default all tools and knowledge base retrievals are hidden for users (same behavior as in previous versions). You can optionally provide a display title and description to be shown to the user.

When applying User Access and Approval, the chat completions (or call agent) operations behave differently than in previous versions, because the tool call response is sent back to the microflow without the final assistant response:

• Action microflows do not need to change, only if pre- or postprocessing was added that should not happen multiple times per sending the user message 
• Tool messages are now persisted in the database and sent along next chat messages. This means an LLM can reason about previous toolcalls that happened in the same chat. Note that this increases the input token consumption which might increase overall costs. This can be set off if the "SaveToolCallHistory" boolean on the request object is set to false before calling the operations. If any tool is set for user visibility or confirmation, this setting gets ignored. 

We changed how we apply custom styling in the module to the snippets and pages to make the styling more compatible with different design systems than Atlas.

This module version is compatible with the GenAI Commons 6.0.0 and AgentCommons 3.0.0 version.

We fixed several small bugs such as that the link to a trace did not work.

We added several enhancements for AI Traces:

* For easier insights into duration and order of AI interactions, a Gantt chart optic is now used.

* Traces for deleted agents and agent versions are greyed out but still contain information such as the version id.

* When filtering traces with a range, a slider can now be used (instead of a text box).

The module was upgraded to the latest LTS version 10.24.13.

A new capability and role have been added to inspect agent and other AI interaction traces in runtime for improved observability and troubleshooting. Users can now better understand system behavior and performance with:

- Detailed visibility into each LLM request and response including tool calls and knowledge retrievals

- Execution metadata (timestamps, duration, etc.).

Note that the AgentCommons module is now a dependency for the ConversationalUI module in order to filter for agents in the UI.

Static image widget instances were replaced with Image widgets.

UI for Mistral Deployed models has been added.