Background Agent Configuration - SubVerse AI

Identity

Field	Type	Default	Description
Name	Text	— (required)	Unique identifier. Lowercase letters, numbers, and underscores only (e.g. `data_processor`). Must start with a letter.
Description	Text	`""`	Internal description of what task this agent performs.
Architecture	Dropdown	`Prompt Based`	Prompt Based uses a system prompt. Conversational Flow uses a visual node-based builder.
Debug Mode	Toggle	Off	Outputs extra internal logs. Useful when building and testing workflows.

Agent Settings

Field	Type	Default	Description
Greeting Message	Text	`""`	The initial prompt or task description sent to the agent when it is invoked.
Greeting Type	Dropdown	`Prompt`	Prompt — agent generates a response based on the context. Say — outputs exact text (rarely used for background agents).
Who Speaks First	Dropdown	`SubVerseAI Agent`	Keep as SubVerseAI Agent so the agent begins processing immediately when triggered.
System Prompt	Text area	Empty	The main system prompt. Describe clearly what input the agent receives, what to do with it, and what output to produce. Use the Refine using AI button for assistance.

Large Language Model Preferences

Field	Type	Default	Description
LLM Model	Dropdown	— (required)	The AI model powering this agent. Choose a capable model for complex tasks. Each model shows its speed and reasoning ratings.
Thinking Level	Dropdown	`Default`	Extended reasoning depth. Options: Default, Low, Medium, High, Very High. For complex tasks like document analysis or data reasoning, use High or Very High.
Max Tokens	Number	None	Maximum tokens the LLM can generate per response. Leave blank to use the model’s default.
Temperature	Slider	`0.5`	Response creativity. For structured tasks (classification, extraction) use `0.0`–`0.3`. For generative tasks (drafting, summarizing) use `0.5`–`0.8`.

Session Settings

Field	Type	Default	Description
Communication Method	Dropdown	`Two Way`	Two Way — agent exchanges messages internally as part of its task. One Way — agent only produces output.
Session Timeout (mins)	Number	None	Minutes after which an idle background task is ended automatically.
Remember Customer Interactions	Toggle	Off	Save a summary of this execution to memory when the task completes.
Use Customer Interaction Memory	Toggle	Off	Pull past execution summaries into context when this agent is triggered again for the same user or entity.

Knowledgebase Settings

Field	Type	Default	Description
Knowledge Base	Dropdown	None	Attach a Knowledge Base. The agent will search it as part of completing its task. Useful for question-answering or document-grounded tasks.

Function Settings

Field	Type	Default	Description
Pre Call Tools	Multi-select	Empty	LLM tools that run before the agent starts its task (e.g. fetch data from an external system).
During Call Tools	Multi-select	Empty	Tools the agent can call while executing (e.g. write to a database, call an external API, send an email).
Post Call Tools	Multi-select	Empty	LLM tools that run after the task completes (e.g. store results, notify another system).

Agent Security Settings

Field	Type	Default	Description
Allow List (Allowed Callers)	Text area	Empty	Regex patterns for sources allowed to trigger this agent, separated by commas. Leave empty to allow all.
Block List (Blocked Callers)	Text area	Empty	Regex patterns for sources that are blocked, separated by commas. Takes priority over Allow List.

Post Call Data Analysis

After each execution, the agent can automatically analyze the output and extract structured data.

Field	Type	Description
LLM Model	Dropdown	The AI model used to run the post-execution analysis. Can differ from the main agent model.
Parameters	Form / JSON	Define the data points to extract from each execution. Switch between Form (add parameters one by one) and JSON (paste a raw JSON schema).

Parameter fields:

Field	Description
Parameter Name	The key name for this data point (e.g. `task_completion`, `summary`).
Details	Description of what this parameter represents and how to extract it.
Type	Data type: `string`, `boolean`, `number`, etc.
Required	Whether this parameter must be present in every analysis output.