Skip to main content

​
What artifacts are

Artifacts are structured work products an agent creates during conversation — documents, spreadsheets, UI layouts, code, images, and more. Unlike plain chat text, artifacts are saved, versioned, and listed in the Artifacts library for your workspace. They are the durable output of agent work: the memo you edit after the chat ends, the table you export, the component you publish to a registry page.
Screenshot placeholder Artifacts library page: table of artifacts with columns for title, type, agent, chat, and last updated. Select a row to preview in the right panel.

​
How artifacts are created

  1. Assign the matching builder Focus to the agent — see Focus & capabilities.
  2. Chat or use the Lab — ask the agent to create the output explicitly (“Draft this as a document,” “Build a comparison table”).
  3. The artifact opens in a side panel beside the conversation. Continue instructing the agent to refine it.
  4. Each meaningful update saves a new artifact version linked to that chat.
If the agent only replies in text, it may lack the right Focus — or you may need a clearer create request.

​
Artifact types

​
The Artifacts library

Open Artifacts from Agent Manager to browse everything saved across chats. Select an artifact to:
  • Read and edit the latest content in the appropriate editor
  • Review version history — each save from the agent or your edits
  • See which agent and chat produced it
Screenshot placeholder Artifact detail view: editor on the left, version timeline on the right. Show version list with timestamps and optional link back to source chat.
Artifacts can be shared via share links so someone outside the original chat can view the output — useful for reviews, client approvals, or handoffs.
Screenshot placeholder Share link control on an artifact detail page and the read-only share viewer page a recipient sees.
Treat share links like any sensitive link — they expose the artifact content to holders of the URL. Use for appropriate audiences only.

​
Editing after chat

You can edit artifacts directly in their editors without returning to chat. The agent’s version history remains; your manual edits create new versions too. To have the agent continue refining, reopen the source chat or start a new conversation with the same agent and reference the artifact.

​
Requesting new artifact types

XSpada adds builder types as product capabilities grow. If your workflow needs an output format that is not listed:
  1. Describe the job — what the agent should produce, who consumes it, and how it differs from existing types.
  2. Check existing types — Text editor and Table editor cover many cases; Registry component builder may already fit registry work.
  3. Contact your XSpada rep or explore options in XSpada Evolve for use-case fit.
Clear examples (“We need agents to produce X that our team then does Y with”) speed evaluation more than generic “we need more artifacts.”

​
Where to go next

Context, files & artifacts

Linked files vs. artifacts vs. chat attachments.

Focus & capabilities

Which Focus unlocks each builder type.
If you would like something unique to your business or have a unique issue, use XSpada Evolve or ask a XSpada rep.