Agents

An agent is one configured AI Shopping Assistant in your organization. You can have several. Each agent keeps its configuration in versions, so you can change how it behaves without losing what came before, and test a change before it goes live.

Open Agents in the dashboard sidebar. If the current site already has at least one agent, this takes you straight into its workspace (no detour through a list): your only agent, the site's configured default agent (set on the site's Public access settings), or, failing that, its most recently created agent. A + icon next to the sidebar item lets you create another agent from anywhere, without opening a list first. It shows up for anyone holding Agents: Edit.

To see every agent on the site instead of being sent to one, or when the site has none yet, go to /dashboard/sites/{siteId}/agents. That page also adapts to how many agents the site has: a create prompt with none, a straight redirect into the workspace with exactly one (the same shortcut the sidebar takes), and the full table below with two or more. To look the same data up from outside the dashboard, see the API reference.

How agents workCopy Anchor Link

• An organization has many agents. Each agent has a type that selects its base persona: Shopping assistant (the built-in shopping persona) or Custom (no built-in persona at all: everything the agent does comes from its version's own instructions).
• An agent's configuration lives in versions. Each version has its own instructions (extra guidance layered on top of the type's base persona) plus which connectors, chat tools, and capabilities it can use. The type also decides a new version's default chat tools: all four for Shopping assistant, none for Custom; capabilities always default off, for every type. See Configuration versions.
• A new agent starts with a v1 champion that has empty instructions, so it works immediately.

Create an agentCopy Anchor Link

The agent detail pageCopy Anchor Link

An agent's detail page is a tabbed workspace: each tab is its own URL, so deep links and refresh work. The agent's name is an inline, editable control in the page header, next to a read-only type badge (the type can't change after creation), visible above every tab, not just one of them. Four tabs are equally-weighted and cover what you're usually doing; Versions and Evaluations sit after a divider, in smaller type, for less frequent needs:

• Overview: the default landing tab. A status card shows whether the agent is live (and since when) or has no champion yet, when the agent was last modified, and (once it has a champion) a "Test it on Playground" shortcut that opens the champion version directly. Below it, quick-link cards into Identity, Capabilities, and Instructions each show that section's real live facts (the identity's display name and avatar status, the engine plus its tool/capability/connector counts, the instruction block count) rather than a bare link (or, with no champion yet, a prompt to create the first version). Two more cards show actual usage: conversations in the last 7 days with a link to the full list, and the latest Evaluations run with its score and a link to it. A compact block of "coming soon" previews (performance, A/B testing, engine proposals) closes out the tab. Which agent answers the site's public link is chosen on the site's Public access page, not here.
• Identity, Capabilities, Instructions: one shared editing surface, not three separate pages: landing on any of these three tabs means editing the agent's current configuration directly; there's no separate "start editing" step. Switching between them never loses what you've typed, since all three share one mounted form. See Configuration versions for what each groups and how saving works.
• Versions: the version registry and its lifecycle actions (promote, archive, duplicate). See Configuration versions.
• Evaluations: this agent's evaluation run history, and a quick-launch entry point to test a version against a library of scenarios. See Evaluation.

There's no Playground tab and no Insights tab here. To test a version against the live model, use its Playground action from the Versions table's ⋯ menu, from a version's own read-only page, or from Overview's shortcut once the agent has a champion. All three open the standalone Playground with that exact version pinned.

Switching between agentsCopy Anchor Link

When the site has two or more agents, opening any one of them shows a selector column beside the tabbed workspace: every agent on the site, by name, plus a New agent action in its own footer. Pick another agent to switch straight into its workspace: no trip back through the agents list. This column isn't shown on a single-agent site (there's nothing to switch to) or on the agents list/table page itself.

Public identityCopy Anchor Link

By default, a shopper sees a generic assistant icon and the label "Assistant" (translated into whatever language the chat is running in), never the agent's internal name, which is an admin-only label (often a technical identifier, not something meant for a shopper).

To give an agent its own face and name, open its Identity tab and use the avatar and display name fields:

• Avatar: upload a PNG, JPEG, or WebP image, up to 2 MB. Shown as a circular avatar. The file uploads as soon as you pick it (you don't wait for it on save). Clear avatar falls back to the generic icon; the uploaded file itself is kept, not deleted, since an earlier configuration version may still reference it.
• Display name: one tab per language the site has enabled, each holding a single name up to 40 characters. A language left blank falls back to the generic "Assistant" label in that language, never to another language's configured name, and never to the agent's internal name.

Both fields are independent and optional. They behave like every other field on the Identity/Capabilities/Instructions editing surface: it lives on a configuration version, not on the agent directly, so nothing is saved until you use the surface's own save bar:

• Save creates a new draft: nothing goes live, even if the agent already has a champion.
• Publish creates a new champion immediately, carrying forward the rest of the source version's configuration (instructions, engine, tools, connectors, conversation starters) unchanged.

The Identity tab always starts from the champion's current identity (or the most recent version's, if there's no champion yet), so on an agent without a champion it starts blank, even if an existing draft already has a name or avatar set. A draft or candidate version can carry a different name and avatar than what's actually live, the same guarantee every other version field already had: nothing a shopper sees changes until that draft is promoted to champion.

Editing these fields requires the Agents: Edit permission, same as the rest of this page. A version created before identity was versioned shows a note on its read-only page that its identity was copied in by a one-time migration rather than configured by a merchant for that version.

Removing a language from the site's enabled languages also deletes that language's display name from every agent on the site, alongside the existing welcome-message cleanup and that language's starter-question text, behind the same confirmation.

Delete an agentCopy Anchor Link

Delete an agent from the row menu in the agents table, at /dashboard/sites/{siteId}/agents. It's a soft delete: the agent disappears from every list, but its versions and conversations are kept: the transcripts you've collected aren't destroyed. The deletion can't be undone from the dashboard.