Agents & How to use them
Fater includes Agents – AI assistants integrated into the Settings Tab of the Right Sidebar. You can chat with an Agent to get help, refine ideas, and even have it assist with certain editor tasks. You can also manage and customize different Agent configurations.
What is an Agent?
An Agent in Fater is essentially a pre-configured AI powered by a Large Language Model (LLM). Each Agent configuration includes:
Name: A recognizable name for the Agent (e.g., "Helpful Assistant," "Creative Prompter").
Description: A brief summary of its intended purpose or specialty.
System Prompt: This is the core instruction set that defines the Agent's behavior, tone, expertise, and how it should respond to your queries. It's like giving the AI its "job description" and "rules of engagement."
LLM Model: The specific underlying language model (e.g., versions of GPT or similar) that powers the Agent's responses.
Temperature: A setting that controls the "creativity" or randomness of the Agent's replies. Higher values lead to more diverse but potentially less predictable answers; lower values make responses more focused and deterministic.
Custom Agents: While Fater provides default Agent(s), you can also create and customize your own Agents to tailor their behavior for specific tasks or a preferred interaction style. (See How to use the assistant for details).
The Role of the System Prompt
The System Prompt is the most crucial part of an Agent's configuration. It's a hidden instruction given to the LLM before your conversation starts. A well-written system prompt can dramatically influence:
The Agent's Persona: Should it be formal, casual, technical, creative?
Its Knowledge Domain: Should it focus on Fater features, general art concepts, or specific industry jargon?
How it Responds: Should it be concise, detailed, ask clarifying questions, or offer suggestions?
Task Capabilities: Can it help draft image prompts, explain parameters, or assist with actions in the editor?
(Example: Currently, Fater includes a default Agent. In the future, we plan to add an Agent with this documentation as its context, allowing it to answer your questions about how to use Fater directly!)
Understanding LLM Models and Temperature
LLM Model: Different LLMs have varying strengths (e.g., some might be better at creative writing, others at technical explanation). Fater allows you to select from available LLMs when configuring an Agent.
Temperature:
Low Temperature (e.g., 0.2 - 0.5): Produces more focused, deterministic, and often more factual responses. Good for tasks requiring precision or consistent answers.
High Temperature (e.g., 0.8 - 1.0+): Encourages more creative, diverse, and sometimes unexpected responses. Good for brainstorming or exploring varied ideas.
How the Agent Interacts with the Editor
The Agent isn't just a chatbot; it can interact with the Fater editor in meaningful ways:
Setting Prompts for Image Editor: You can ask the Agent to help you write or refine a prompt for the Image Editor. If the Agent provides a suitable prompt in its response, you can often use a context menu action (right-click on the message) to "Set as prompt," which will copy that text directly into the Floating Prompt Area.
(More advanced Agent-driven actions and editor integrations will be introduced in future updates.)
Using the Agent
Open the Settings Tab in the Right Sidebar.
Select an Agent from the dropdown menu.
Chat with the Agent by typing your questions or requests into the chat input area and pressing Send.
Review its responses and utilize any suggested prompts or information.
The Agent is a valuable tool for learning Fater, brainstorming ideas, and streamlining parts of your creative workflow.
Last updated