How it works

Your agent does the work. You make the call.

Connect the agent you already use, give it what it needs to know, and review what it prepares. Three steps, and the last one is always yours.

Three steps

From connected to decided.

1

Connect your agent

Run npx xevol connect claude, or codex or pi. Approve with a code in any browser and choose one Workspace.

Xevol Settings listing three connected AI agents, Claude Code, Codex and pi, each with a disconnect button.
2

Give it context

Save the links, videos, notes and files that matter. Your agent works from the same sources and records you do, and answers cite where they came from.

An Xevol answer to “Why do teams stop using a new tool?” citing two saved sources.
3

Review what it drafts

Outreach, Offers, records and research arrive as drafts with their evidence. You approve or reject each one.

An outreach draft in Xevol marked Needs review, with its evidence, an evidence check, and Approve draft and Reject draft buttons.

The rule

Agents prepare. People decide.

Your agent can

  • Read every record in its Workspace
  • Create and update drafts
  • Start research and enrichment
  • Submit work for your review

Only you can

  • Approve or reject a draft
  • Send, publish or delete
  • Change billing, members or modules
  • Disconnect the agent

Context

What your agent can work from.

  • Articles
  • Videos
  • Interviews
  • Notes and files
  • Contacts
  • Companies
  • Deals
  • Leads
  • Offers
  • Structs
  • Workspace skills

Connect your first agent today.

Free to try. No card required.