| Term | Meaning |
|---|---|
| Ask mode / Auto mode | The run mode dropdown in the prompt box. Auto mode (default) automates decisions and shows you only the final result. Ask mode has you review each step before proceeding. See Running a Task. |
| Connector | A real, OAuth-authenticated link to a third-party app (Slack, Notion, Gmail, and 30+ more) that lets the agent read and act on your actual data. See Connectors. |
| Customize | The sidebar section holding Connectors, Sources, Memory, and Personalization. |
| Memory | A toggle plus an auto-extracted list of facts the agent remembers across chats. |
| Model tier | Flash, Pro, or Ultra — speed vs. depth. Chosen from the dropdown next to + in the prompt box. |
| Personality | A custom persona built in Personalization — name, profession, traits, and background — that shapes how the agent responds. |
| Plan | The numbered, editable step-by-step outline the agent proposes before running a task. |
| Scheduled task | A recurring agent job, configured with a trigger (repeat frequency + time), output format, and output destination. |
| Startup | Imagine Computer’s website builder for games, 2D or 3D sites, and landing pages. It has its own Plan/Build mode and Eco/Balanced/Max tier controls. See Startup. |
| Source | A named, uploaded collection of your own files (documents or code) available to reference in any chat. |
| Tool | A card under Get started with Tools on the home. Each is a preset on the same chat framework (Deep Research, Plan Trips, Create Slides, etc.) — not a separate tool, with one exception: Startup. |
| Tools permission | The per-connector, per-action allow/deny list controlling exactly what a connector can do. |
Glossary
A quick reference to terms used throughout this guide and inside Imagine Computer.
A quick reference to terms used throughout this guide and inside Imagine Computer.