# Terminology Aubrey uses a few words that may be new at first. This page explains them in plain language. You do not need to memorize these terms before using Aubrey. If you can describe what you want help with, Aubrey can usually guide you from there. ## The most important words ### Aubrey Aubrey is the product. It is the personal AI agent service you use from the web, email, calendar, and supported messaging channels. ### Agent An agent is the AI helper you create inside Aubrey. If you have used other AI products, you may have heard this called an AI assistant. In Aubrey, we use the word agent. We use "agent" because Aubrey is meant to help with tasks, not just answer one question at a time. In the current AI world, an agent usually means AI that can understand a goal, use the information and connections you approve, and help move the work forward. For example, an Aubrey agent can help summarize emails, prepare a plan, remember useful preferences, create reminders, or coordinate with other people. It is still under your control: you choose what to connect, what to ask, and what to approve. You can have one agent for general everyday help, or create separate agents for different areas of life. For example: * a family planning agent * an email follow-up agent * a project momentum agent * a research and writing agent ### Workspace The workspace is the agent's working area in the web app. It is where you can chat with the agent and manage things connected to that agent, such as recipes, skills, channels, sharing, email, calendar, and history. ## How Aubrey gets things done ### Recipe A recipe is a ready-made routine you can turn on for an agent. Recipes are meant to help with common repeated work. For example, a recipe might help Aubrey summarize important emails, prepare a planning brief, or remind you about follow-ups. Think of a recipe as: "Do this kind of helpful routine for me." ### Skill A skill is a reusable way Aubrey knows how to help. Some skills are built into Aubrey. Some are connected to services you choose. Some can be created from patterns Aubrey learns over time. You do not need to install code or manage technical settings to use skills. In normal use, skills simply make Aubrey better at a kind of task. ### Integration An integration is a connection between Aubrey and another service you choose. For example, Google Workspace can let Aubrey work with Gmail, Calendar, Drive, Docs, Sheets, and selected files when that feature is available on your plan. You stay in control of what you connect. If you disconnect an integration, Aubrey should stop using that connection. ### Channel A channel is a place where you can talk with Aubrey or where Aubrey can send updates. Common examples include: * web chat * Aubrey email * WhatsApp or Telegram when available for your account * calendar reminders or calendar feeds Think of channels as "where the conversation happens." ## Useful terms you may see in the app ### Aubrey email address Each agent can have its own Aubrey email address. You can forward selected emails to that address, email Aubrey directly, or cc Aubrey in a thread with a clear request. ### Managed calendar A managed calendar is a calendar Aubrey can use for reminders, holds, check-ins, appointments, and follow-ups. It is separate from a normal chat message. It gives Aubrey a place to put time-based items. ### Calendar feed A calendar feed is a private link that lets you view Aubrey calendar items in another calendar app. Treat calendar feed links like private links. Anyone with the link may be able to see the calendar items it exposes. ### Memory Memory is information Aubrey can remember for later. For example, Aubrey might remember that you prefer short summaries, that school pickup is usually at a certain time, or that a family member has soccer on Tuesdays. Use memory for details you want Aubrey to reuse. Avoid adding information you do not want Aubrey to keep. ### Context Context is the background information Aubrey uses to answer or act. Context can include the current chat, forwarded emails, selected files, calendar items, lists, preferences, or details you have asked Aubrey to remember. ### Files Files are documents or attachments you give Aubrey so it can use them for a task. For Google Workspace, selected files are files you choose for Aubrey to access through Google Drive, Docs, or Sheets when that connection is available. ### Sharing Sharing lets another person use the same agent or see a shared item. Use sharing when a task belongs to more than one person, such as family planning, a shared event checklist, or a group chat. ### Collaborator A collaborator is someone you invite to use or contribute to an agent, shared list, message thread, or connected channel. Only invite people who should be able to see that shared context. ### Approval An approval is a moment where Aubrey asks you to confirm something before it continues. You might see approvals for decisions, memory changes, or actions that need your review. ### Run A run is one attempt by Aubrey to handle a request. You usually do not need this word. If a page says a run failed or can be retried, it means Aubrey had trouble completing that request and you can ask it to try again. ## Terms you should not need often Some words are mostly internal to Aubrey. You may see them in an error message or support conversation: * **Tenant:** your account or workspace boundary behind the scenes. * **Provider:** an outside service Aubrey connects to, such as Google. * **Credential:** the secure connection information Aubrey uses after you approve an integration. If you see one of these words and are not sure what it means, contact support with a screenshot or the exact message. --- # Agent Instructions: Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://aubrey-agent.gitbook.io/aubrey-docs/documentation/getting-started/terminology.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.