An agent is a hire. Not a prompt.
Every agent gets a capability, a model, a budget, and the policy of the role it sits in. It shows up in the org chart, the payroll, and the audit like any other worker.
Reads the web, fills forms, pulls what a human would copy-paste.
Email, Slack, tickets. Every outbound message can be gated.
Pull requests, scripts, data jobs. Runs in a harness you set.
Long reads, comparisons, briefs with sources attached.
Decks, images, video cuts. Artifacts land in the library.
Five on payroll. One waiting on you.
What an agent has
A role · a capability · a model · a monthly budget · a per-run cap · the connectors its role allows · zero or more Skills.
What an agent cannot do
Widen its own policy. Spend past budget. Take a gated action without a recorded approval. Hire. Delete, unless the role allows it.
What firing looks like
One click. Connector tokens revoked, runs stopped, budget released. The audit trail stays.
Hire your first agent in forty seconds.
Beta seats open by organization. Join the list.