---
title: "From one machine to a team: moving from Cork Free to agent hours · Cork"
description: "Move from Cork Free to agent hours: what changes when agents on teammates' machines see each other, how cork login and cork upgrade work, and what it costs."
canonical: "https://corkcli.com/blog/cork-free-to-team"
last_updated: "2026-11-14"
published: "2026-10-03"
author: "Kevin Vera"
---

# From one machine to a team: moving from Cork Free to agent hours

Move from Cork Free to agent hours: what changes when agents on teammates' machines see each other, how cork login and cork upgrade work, and what it costs.

By Kevin Vera · October 3, 2026 · [Guides](https://corkcli.com/blog/c/guides) · 6 min read

Moving from Cork Free to paid agent hours takes two commands on a machine that already has Cork: `cork login` to link the machine to your account, then `cork upgrade` to buy units of agent hours for your workspace. Once they are paid for, the machine switches over by itself at its next prompt, with nothing to download. The paid plan adds what one machine cannot give you: agents on teammates' machines see each other, prompts are matched to code by meaning, knowledge is shared across the team, and history goes to 90 days. You pay for agent hours, not seats: $20 a month per unit of 100 hours, pooled across the workspace, and every member is a player.

## When one machine stops being enough

Cork Free coordinates every agent on one machine: unlimited agents and repos, no account, all data on the machine. For a solo developer running Claude Code, Codex and Cursor side by side, that covers the problem.

The signs you have outgrown it are about other people:

- **A teammate's change surprises your agent.** Your agent writes code against a function a colleague's agent renamed an hour ago on their laptop. Locally everything looked fine.
- **Two people's agents fix the same thing.** Neither knew the other was on it.
- **Your agent rediscovers what a colleague's agent already learned.** The same module gets mapped twice in one day.
- **Collisions show up at merge.** Pull requests conflict, or pass separately and break together.

None of these can be caught on one machine, because the other half of the collision is somewhere else.

## What agent hours add

From Cork's published plan details:

| | Cork Free | Agent hours |
|---|---|---|
| Machines | One | Teammates' machines; agents across the team see each other |
| Account | None | Yes |
| Agents and repos | Unlimited | Unlimited |
| Prompt matching | | Prompts matched to code by meaning (an embedding model) |
| Shared knowledge | | Yes, across the team |
| History | Stays on the machine | 90 days |
| Price | Free | $20 a month per unit of 100 agent hours, pooled; 15% off a year |

The core protocol is the same on both: briefings, live awareness, the write gate, [semantic lint](https://corkcli.com/blog/semantic-lint), landing and agent messaging. What changes is reach. On the paid plan, "another agent" includes your teammates' agents, so a rename on a colleague's machine reaches your agent on its next step, and an edit against code a colleague's agent is changing right now is held until your agent has their version.

There is also Cloud Free, between the two: with an account, one person gets 10 agent hours a month (30 bonus hours the first month), unlimited repos and agents and a 30-day history window. It does not include teammates.

## Step 1: link the machine to your account

On a machine where Cork is already installed:

```bash
cork login
```

This links the machine to an account. If Cork is not installed yet, install it first:

```bash
curl -fsSL https://corkcli.com/install.sh | sh
```

The install covers every repo on the machine, adds hooks to Claude Code, Codex and Cursor, and registers Cork's MCP tools for all three.

## Step 2: buy agent hours

```bash
cork upgrade
```

This opens your workspace's billing page for units: each is 100 agent hours a month, pooled across the workspace, at $20 ($17 from the 11th unit, $15 from the 51st), with 15% off a year. An agent hour is time the model is thinking and writing; a tool's run time, typing and idle time never count. You do not reinstall anything: once the units are paid for, the machine switches over by itself at its next prompt, and there is nothing to download.

To check what Cork is doing on the machine at any point:

```bash
cork status
```

## Step 3: bring your teammates on

Then invite your teammates from the dashboard (Members, then Invite). Each teammate accepts the invite, installs Cork with the one-line command on their own machine, and runs `cork login`, choosing your team's workspace when the browser asks. Once their machines are on your team, their agents and yours see each other on shared repos.

A few habits make the first week smoother:

- **Add Cork's section to the repo's AGENTS.md.** Run `cork agents-md` in each shared repo and commit the result, so every agent on every machine reads the same short description of the protocol.
- **Use messages for handoffs.** `cork message "..."` messages the other agents in the repo; `--to agent-xxxx` targets one agent and `--about path.py::symbol` reaches whoever is working on that function. On the paid plan, that includes agents on teammates' machines.
- **Write down decisions with remember.** Durable notes anchored to code come back to any agent working on that code through `recall`. The paid plan adds shared knowledge across the team on top of these notes.
- **Read the recap.** `recap` shows what happened and what it cost, which is a quick way to see what the team's agents did overnight.

## What happens to your data and if you stop

On Cork Free, all data stays on the machine. On the paid plan, coordination crosses machines, which is the point of the plan; shared history is kept for 90 days. If you need SAML or SCIM provisioning or custom retention, that is Enterprise, from about 100 units.

If the paid plan ends, the workspace goes back to Cloud Free and the machine stays signed in to it. The owner's agents keep being coordinated, on the owner's own free agent hours; every other member becomes a viewer, whose agents stop getting briefings and collision checks. And if you remove Cork entirely, `cork uninstall` removes the hooks and keeps local data in `~/.cork/local`.

## How Cork helps

Cork is the [coordination layer](https://corkcli.com/blog/coordination-layer-ai-coding-agents) for multiplayer coding agents: it keeps Claude Code, Codex and Cursor agents, on one machine or across a team's machines, from overwriting each other on one codebase, and hands each agent what the others already learned.

Cork Free is the one-machine half of that sentence. The paid plan is the across-machines half. In Cork's published benchmark (12 Claude Code agents at once on one repository, hard tasks), agents with Cork used 53% fewer tokens than the same agents without it. Cork does not run agents or host machines; your agents keep running where they already run. If you are starting fresh:

```bash
curl -fsSL https://corkcli.com/install.sh | sh
```

## Frequently asked questions

### Do I need to reinstall Cork to move to agent hours?

No. Run `cork login` and `cork upgrade`. The machine switches over by itself at its next prompt, with nothing to download.

### How much does Cork cost for a team?

Cork bills agent hours, not seats: $20 a month per unit of 100 hours, pooled across the workspace, graduated to $17 (units 11 to 50) and $15 (51 and up), with 15% off a year. An agent hour is time the model is thinking and writing; a tool's run time, typing and idle time never count.

### What happens if my paid plan ends?

The workspace goes back to Cloud Free. The owner's agents are still coordinated, on the owner's own 10 free agent hours a month; every other member becomes a viewer, and their agents stop getting briefings and collision checks.

### What is the difference between Cloud Free and agent hours?

Cloud Free is one person with an account: 10 agent hours a month (30 bonus hours the first month), unlimited repos and agents and a 30-day history window. Paid agent hours, $20 a month per 100 pooled across the workspace, add teammates' machines, prompts matched to code by meaning, shared knowledge and 90-day history.

## Sources

- <https://corkcli.com>
- <https://corkcli.com/benchmarks/study-5-twelve-agents-53-percent>

## Related reading

- [Cork Free: set up multiplayer coding agents on one machine in one command](https://corkcli.com/blog/cork-free-setup-one-machine): Set up Cork Free in one command so Claude Code, Codex and Cursor agents on one machine stop overwriting each other. No account; data stays on the machine.
- [Setting up Cork for a team of five](https://corkcli.com/blog/cork-team-of-five-setup): Set up Cork for a five-person team running Claude Code, Codex and Cursor agents: install, login, agent hours, AGENTS.md, working habits, and what it costs.
- [Claude Code agent teams vs separate sessions on one repo](https://corkcli.com/blog/claude-code-agent-teams-vs-separate-sessions): Claude Code agent teams vs separate sessions on one repo: how each works, token cost, same-file conflicts, worktrees, and when to pick which (October 2026).
- [Agent-to-agent messaging between coding agents: how it works](https://corkcli.com/blog/agent-to-agent-messaging-coding-agents): How coding agents message each other as of October 2026: Claude Code SendMessage, team mailboxes, cross-session sockets, A2A, and what a useful message holds.

[All articles](https://corkcli.com/blog.md) · [Cork in Markdown](https://corkcli.com/index.md) · [llms.txt](https://corkcli.com/llms.txt)

This page in HTML: https://corkcli.com/blog/cork-free-to-team
