# One operator seat for agents wherever they run.

> Start beside your local toolchain, move an interactive session to cloud capacity, or let CI wake a tightly scoped Factory agent. Hive keeps the session, branch, PR, CI and spend in one workspace.

Source: https://hiveci.io/agents
Language: en
Alternate: https://hiveci.io/de/agents

The lanes share a control plane—not one undifferentiated security model.
| | |
|---|---|
| WORKSTATION | real worktree · local tools |
| CLOUD | ephemeral pod · interactive |
| FACTORY | CI task · reviewed wrapper |
## Choose the boundary that fits the task.
Hive does not pretend a laptop shell and a CI container are the same thing. It gives each lane the controls and containment its job requires.
- **Workstation: Work where the repository already lives** — A launched local agent gets a real git worktree, the repository setup and explicitly allowed build caches. Source stays on that workstation unless you choose a workflow that uploads it.
- **Cloud: Free the laptop without losing the session** — An interactive Pi session runs in an ephemeral, resource-capped pod. Steer it from the browser, then let the delivery wrapper preserve its commits and PR.
- **Factory: Let CI open a bounded coding task** — A pinned Pi runtime runs as an ordinary Hive workload. Compiled context, reviewed capabilities, budgets and git policy stay outside the model.
- **Workspace: Operate the fleet from one place** — Live transcript, plan, workflow, files, branch, PR, CI and approvals sit beside each session instead of across terminal tabs.
- **Teams: Coordinate work without hiding collisions** — Group sessions into a team, see branches and claims together, pass durable notes, and surface overlapping work before delivery.
- **Ownership: Controls follow the person and the lane** — Conversations are owner-only by default. Viewer shares are read-only; drive access is reserved for cloud sessions, where the owner’s machine is not exposed.
## The same session from a terminal, from MCP and from the browser.
A session is not a tab you have to keep open. It is addressable.
hive launch starts a session on the machine you name and attaches you to it. From then on the session reports its phase, plan, worktree, branch and pull request into the workspace, whether or not anyone is watching.
steer_agent queues a message the session claims on its next turn; read_inbox is how a session hears from its team. Both are MCP calls, so the agent you already use can drive them without opening Hive at all.
```
$ hive launch --repo acme/shop --branch fix/checkout-total --target workstation
session as-7f3c · workstation · worktree ready

steer_agent { "session_id": "as-7f3c", "text": "Keep the fix inside orders/." }
→ queued · claimed by the session on its next turn

read_inbox { "session_id": "as-7f3c" }
→ 1 team_message: "PR #4822 is green — add a changelog line before merge."
```
## The agent does not sit beside CI. It is inside the same control loop.
Hive already knows the run, tests, branch and delivery state. Agents receive narrow calls to that system instead of credentials to scrape it.
- **Sessions stream into the browser.** — Follow turns, tool calls, costs, plans, workflows and attention state without attaching to the machine that hosts the agent.
- **Steer, interrupt and approve.** — Owner controls ride a durable command channel. A queued command remains visible until the client claims and reports it.
- **The branch and CI tell the truth.** — Working-tree state, pull request and Hive checks share the rail, so “done” is measured against delivery rather than the last assistant message.
## Sharing is explicit, and write access is narrower than read access.
The UI explains which controls are possible instead of rendering every session as equally steerable.
| Role | Controls | Available on |
|---|---|---|
| Owner | Read · steer · interrupt · approve · end | Every own session |
| Viewer | Read transcript and live stream | Named read-only share |
| Driver | Steer · interrupt · switch model | Cloud sessions only |
| Factory observer | Run, transcript, result and spend | No workstation control |
## Teams already running coding agents, who want them beside the branch.
| | |
|---|---|
| You already use | Claude Code, Cursor or Codex, with a subscription or key you pay for. Hive authenticates its agents as you and never marks up a token. |
| You want | Every session visible next to its branch, pull request and CI state — and steerable from the browser when it wanders. |
| This is not | A hosted IDE, a chat product, or a place to write code by hand. The workstation lane is your own machine; Hive is the workspace over it. |
## See the harness and the platform that make these lanes one system.
- **The interactive harness** — Context, tools, plans, tasks, delegation, gates and Hive telemetry around Pi.
- **The control plane and fleet** — Pipelines, scheduling, runners, caches, test databases, APIs and operating boundaries.
## Bring one repository. See the whole system work.
Start on your own runner, keep your existing CI beside it, and watch Hive turn the next eligible red check into a reviewable fix.