Installed Paperclip, Now What !?

Installed Paperclip, Now What !?

Nick Spisak’s follow-up to his viral Paperclip post walks readers through turning a demo into a functioning, tiny AI company, and the thread is worth a look: https://x.com/NickSpisak_/status/2034635430700679445

In short, the setup uses three free, open source pieces that play different roles. Paperclip is the company dashboard, where missions, budgets, and org charts live. gstack is the engineering team, providing 15 specialist skills (planning, QA, shipping, and so on). autoresearch is the R&D lab, able to run many automated experiments overnight (yes, dozens per hour).

The tutorial claims the whole stack installs in under 10 minutes, and then the magic is orchestration. Paperclip assigns a task, autoresearch runs experiments to find the best approach, gstack builds and tests the winner, and Paperclip logs costs and outcomes. It’s a neat division of labor: one tool assigns, one learns, one builds, and one reports. The post includes practical setup snippets and starter prompts, for example asking gstack’s /office-hours to turn an idea into three implementation plans, or telling autoresearch to run 50 iterations and log results.

A few realistic asides appear in the writeup, like the suggestion to start with a single agent and let hiring scale naturally (don’t overdo it day one), and safety tips such as using gstack’s /careful mode when production code is at stake. The repos are linked and everything is open source, so it’s easy to poke around and experiment.

For those curious about structured, small-scale AI teams, this follow-up is a pragmatic next step after a demo. The author also shares a community waitlist for deeper breakdowns: https://return-my-time.kit.com/1bd2720397

Overall, it’s a compact playbook for turning imaginative demos into repeatable systems, and it points toward tooling that could make solo makers surprisingly productive.