Chat panel: Difference between revisions

From Tygron Support wiki
Jump to navigation Jump to search
Line 19: Line 19:
{{:AI Agent#Chat Agents}}
{{:AI Agent#Chat Agents}}
* Textfields and checkboxes to adjust the following Agent Attributes:
* Textfields and checkboxes to adjust the following Agent Attributes:
: {{Agent attributes|suppresscategory=true|allowselflink=true|
*{{Agent attributes|suppresscategory=true|allowselflink=true|
{{:temperature (Agent Attribute)}}
{{:temperature (Agent Attribute)}}
{{:thinking budget tokens (Agent Attribute)}}
{{:thinking budget tokens (Agent Attribute)}}

Revision as of 15:10, 7 October 2026

The Chat Panel is a pop-out panel that lets users interact with AI Agents through Chat Channels, by sending and receiving Chat Messages.

The panel consists of an overview of Chat Channels in your project, as well as an overview (int HTML) of Chat Messages related to the channel. New Chat Messages can be send by typing the message and pressing the Enter key or the Send button. To add new lines to your messages, use the key combination Ctrl+Enter.

Users can also send additional files with their Chat Message by dragging the file onto the text area of the Chat Message. This can be used to:

  • Provide a data file to analyze
  • Provide an image file to describe

Once send, an AI Agent is initialized to process the message. The response of an AI Agent is a continuously updating Chat Message, consisting of two visible parts, Thinking and Answer, shown in order. The thinking part can be toggled.

Settings

The Settings Panel of a Chat Channel can be opened by selecting the Show Setting checkbox. Inside the Settings Panel you can adjust:

  • The Name of the Chat Channel
  • The Instructions of the AI Agent
  • The Skill link opens the Skill Overview panel to manage the available Skills in your project.
  • The Agent Tools available to AI Agents
  • A choice box with the following Chat Agents:
  • WIKI Assistant: Searching the Tygron Wiki for information on features, configurations and guides.
  • Project Coder: Can edit data in your Project, by executing TQL queries or interact with project API endpoints.