> ## Documentation Index > Fetch the complete documentation index at: https://wiredesk.ai/docs/llms.txt > Use this file to discover all available pages before exploring further. # Glossary > What the words used across WireDesk mean, with a link to the page that covers each one. ### Action Something the agent can do beyond answering, such as capturing a lead, booking a meeting, calling your API, transferring a call or asking for a person. Set on the agent's **Actions** section. See [actions](/docs/actions/overview). ### Activity log The last 50 changes anyone made to the workspace, with who made each one. It records changes, not views. See [settings](/docs/account/settings#activity). ### Admin A role that can build and change agents, knowledge, channels and numbers, and work every conversation. It cannot see billing or manage the team. See [team and roles](/docs/account/team-and-roles). ### Agent 1. The AI assistant you build. It answers on website chat, phone and email, from its own knowledge, persona and actions. See [agent overview](/docs/agents/overview). 2. A teammate role that can read and reply to conversations and work the Inbox, but cannot change how the AI agent works. See [team and roles](/docs/account/team-and-roles). ### Allowance (included usage) The usage your plan includes each month, counted at list rates. Replies and voice minutes both draw on it, and it resets on the first of the month. See [billing](/docs/account/billing). ### Allowed domains The sites your widget may run on. Once set, the widget will not load anywhere else. See [allowed domains](/docs/widget/allowed-domains). ### API credential A key saved under **API keys** on the agent's **Actions** section and used by an API call action. It is stored encrypted and never shown to the agent. See [API calls](/docs/actions/api-calls#api-credentials). ### Browser voice A voice conversation with your agent in the browser, used to test it. It is billed like a phone call. See [test your agent](/docs/agents/test-your-agent). ### Business hours The times when a chat or email visitor may ask for a person. Outside them, the agent takes a message instead. See [business hours](/docs/actions/business-hours). ### Channel A way customers reach the agent: website chat, phone or email. Each can be switched on or off per agent. See [channels](/docs/concepts/channels). ### Contact One person, and the conversations they have had with your agents. See [contacts](/docs/workspace/contacts). ### Conversation One exchange between a customer and your agent on one channel, with its full transcript. See [conversations](/docs/workspace/conversations). ### Conversation outcome How a finished conversation ended: **Resolved** (the agent answered what they came for), **Handed to you** (a person took over), **Message taken** (the agent could not help and took a message), or **Abandoned** (they left before the agent could help). See [analytics](/docs/workspace/analytics). ### Extra fields Details you asked the agent to collect during conversations, such as an order number or a preferred time. They are listed in **Inbox → Extra fields**. See [collect data](/docs/actions/collect-data). ### Greeting The first thing the agent says, word for word, on every channel. See [name and greeting](/docs/persona/name-and-greeting). ### Handover Passing a conversation from the agent to a person. On the phone this is a live transfer. In chat and email the conversation is flagged, a teammate replies from the dashboard, and the agent stays quiet until it is handed back. See [handover](/docs/concepts/handover). ### Inbox Where the leads and extra fields the agent collected wait for a person. See [inbox](/docs/workspace/inbox). ### Instructions What the agent is told about your business and how to behave, on every channel. See [instructions](/docs/persona/instructions). ### Knowledge source Something the agent answers from: a crawled website, a note you pasted, or an uploaded file. See [knowledge](/docs/knowledge/overview). ### Lead A name and contact details for a callback, taken by the **Capture leads** action. Leads appear in **Inbox → Leads**. See [capture leads](/docs/actions/capture-leads). ### Owner The person who created the workspace. There is one per workspace. Only the owner can manage the team, billing and the spend cap, or close the account. See [team and roles](/docs/account/team-and-roles). ### Playground The agent section where you try the agent live and change how the chat widget looks. See [test your agent](/docs/agents/test-your-agent). ### Spend cap A hard monthly limit on what you pay beyond your plan's included usage. At the cap, the agent stops answering billable conversations. It starts at \$50. See [spend cap](/docs/account/spend-cap). ### Transfer Handing a live phone call to a person by ringing your transfer number. Set up with the **Hand calls to you** action and the **Transfer number** on **Channels › Phone**. See [hand calls to you](/docs/actions/hand-calls-to-you). ### Usage event One billable unit, recorded as `chat_message`, `email_message` or `voice_seconds`. These make up your month's usage and what counts against the spend cap. See [spend cap](/docs/account/spend-cap#how-usage-is-counted). ### Verified visitor A signed-in user of your site whose identity your server has signed. The agent can then trust who they are. See [verified visitors](/docs/widget/verified-visitors). ### Widget The chat bubble and panel that `widget.js` adds to your site. See [install the widget](/docs/widget/install). ### Widget key The public key, starting `wk_`, that ties the widget snippet to one agent. It is visible in your page source, so protect it with allowed domains. See [install the widget](/docs/widget/install). ### Wording per channel Extra instructions for one channel only, such as shorter answers on the phone. See [channel wording](/docs/persona/channel-wording). ### Workspace Your company's account in WireDesk. It holds your agents, conversations, contacts, team, plan and spend cap. See [team and roles](/docs/account/team-and-roles).