# 3.A AI Actions

These actions allow you to integrate the new AI tools into your processes, enabling their use to be automated.

# 3.A.3 Call Agent

This action allows you to send a request to an agent that has been previously configured in Settings > AI Tools > Agents.

Once one of the available agents has been selected, all the elements configured in Settings > AI Tools > Agents are loaded automatically and can be customized.

The agent is able to perform tasks automatically based on the enabled tools and the knowledge sources available through RAG and LLM.

The System Prompt can be modified, while operational instructions must be entered in the User Message field.

The Output Fields section already includes the response content field by default, allowing the agent's output to be used in subsequent process steps.

Below is an example where the agent is requested to find a solution to a customer support request received through a ticket, using the user\_manual tool, and write the solution as a comment.

Figure 1

<p class="callout info">Additional information is added to the prompt written by the user to provide the context of the request, such as the entities involved in the process, including their name, module, and CRM ID (crmid).  
This allows the agent to understand which CRM entity the request refers to.  
If needed, process variables can be used to make the request even clearer.  
Example: Add a comment to ticket $15-crmid suggesting a solution...</p>

# 3.A.4 Call Process Tool

This action allows you to execute a process tool.

Once the desired tool has been selected, the expected Input Parameters are loaded and can be configured using process variables, as usual.

Unlike sub-processes, this action does not have the same limitations and can be called multiple times within the same process and also from different processes.

Below is an example of a tool call used to add a comment to a ticket. The parameters are those defined in the process tool, which in this case are the ticket ID and the comment text.