Paperclip vs win.sh: Two Operating Models
Paperclip centers an open-source control plane modeled as an agent company. win.sh runs a 24/7 business loop inside an authority matrix.
Founder, Task Machine
Some founders want to design an agent organization: a CEO agent, a CTO agent, goals that roll down into tasks, and a budget for each role. Others want the business watched and handled without drawing any chart. Paperclip and win.sh serve those two impulses, and the gap between them is wider than their shared autonomy language suggests.
Paperclip is a free, MIT-licensed control plane for teams of AI agents. You set up an org chart of roles, goals, and assigned tasks, give agents heartbeats and budgets, and govern them with approvals and audit logs, using whichever agent workers you bring. win.sh runs a hosted 24/7 loop over the accounts you connect and acts inside an authority matrix you set.
How do Paperclip and win.sh compare?
| Dimension | Paperclip | win.sh |
|---|---|---|
| Core abstraction | Tasks assigned inside an org chart of agent roles and goals | A 24/7 loop that monitors connected business accounts and acts |
| Human role | Design the org, set budgets, approve, and audit | Set rules per type of work, then review decisions and receipts |
| Execution | Self-hosted, local or remote, set up with one command | Cloud sandboxes connected to accounts the customer owns |
| Strongest fit | Legible agent roles, governance, and experiments with agent companies | Founders who want the business watched with little daily involvement |
| Recurring work | Heartbeats bring agents back to their assigned goals | Runs continuously, with recurring workflows and a morning brief |
| Control | Governance, approvals, budgets, audit logs, and bring-your-own workers | Approval gates, authority levels, a hard budget cap, and receipts |
| Main cost | The company simulation comes first, and you host and maintain it | The loop acts before the brief, so much of the review happens afterwards |
| Commercial model | Free and MIT-licensed, with a waitlist for a Paperclip app | Self-serve monthly budget from $50 to $10,000, no revenue share |
Prices and features were checked against each vendor's own pages on 6 October 2026.
Who does Paperclip fit?
Paperclip fits technical builders who enjoy the agent-company idea itself and want to run it on their own machines. The org chart makes roles legible, budgets limit each agent's spend, approvals and audit logs record governance, and plugins and skills extend what agents can do. It costs nothing to try, and one command starts it.
The simulated company sits at the center of the model. A repeatable business process, such as a weekly invoice run with a fixed review step, has to be expressed through roles and goals, and the hosting, upgrades, and uptime are yours while its app is still on a waitlist.
Who does win.sh fit?
win.sh fits a founder who wants a working loop without designing an organization. It watches accounts such as Stripe or Plausible around the clock, proposes or takes the next step, and reports each morning through a Telegram brief and the Decisions tab.
You set trust per type of work, and approvals, edits, and rejections become rules. Approval gates hold spending, outreach, publishing, and sensitive changes, a hard monthly cap limits spend, and win.sh takes no revenue share. It is hosted, so there is nothing to run, and its control plane stays closed to the kind of inspection and modification Paperclip's open code allows.
Where does each product put human control?
Paperclip puts control in governance you design. The org chart says who does what, budgets cap each role, approvals stop the actions you choose, and audit logs record the result, all on infrastructure you control.
win.sh puts control in standing policy you tune from its interface, where the authority matrix plays the part of Paperclip's org chart and decides what each type of work may do alone. A useful test for either product is an agent that exceeds its brief, such as a marketing agent that starts editing prices, and how quickly a person would notice.
Where does Task Machine fit?
Task Machine starts from the work instead of the org chart. People and agents share projects and tasks as one team, and each recurring job runs as an explicit workflow with approval, question, and verifier steps. You decide in Chat, approve in one Inbox, and steer in Tasks, and the process stays readable however the agents are organized.
Task Machine is a hosted product, so there is nothing to install, upgrade, or keep online, and it fits operators who want recurring business work run with a person approving the consequential steps. Builders who want a free, self-hosted system to experiment with agent organizations should pick Paperclip.
Which should you choose?
Choose Paperclip if you want to design and govern an agent organization on your own infrastructure at no cost. Choose win.sh if you want connected business accounts watched around the clock without building anything. Choose Task Machine if you want recurring work run as explicit workflows by a team of people and agents, with approvals in one Inbox.