Paperclip is a free, open-source tool that lets you manage a team of AI agents - like having a staff of assistants that research, write, and execute tasks on autopilot. This checklist walks you through every step from installation to your first working agent. No coding experience needed.
| # | Step | Why you need it | What to do | Category |
|---|---|---|---|---|
| A1 | Install Node.js Prerequisites |
Paperclip runs on Node.js - a free program that lets your computer execute modern tools. Without it, the install command won't work. | Go to nodejs.org, download the latest version, and run the installer. To confirm it worked, open your terminal and type: node --version |
Prereq |
| A2 | Install pnpm Prerequisites |
pnpm is a package manager - think of it as an app store for developer tools. Paperclip needs it to download its components. | In your terminal, paste this and press Enter: npm install -g pnpm |
Prereq |
| A3 | Get an API key Prerequisites |
An API key is like a password that lets Paperclip talk to Claude (the AI). Your agents need it to think and work. Costs per use | Go to console.anthropic.com, create a free account, then generate an API key. Copy it somewhere safe - you'll paste it during setup. You pay only for what your agents use (a few cents per task). | Prereq |
| A4 | Open a terminal Prerequisites |
The terminal is where you type commands to install and run Paperclip. It looks like a text-only screen - no buttons, just typing. Sounds scary, but you only need it for 2 minutes. | Mac: Open Spotlight (Cmd+Space), type "Terminal", press Enter. Windows: Press Win+R, type "cmd", press Enter. Or use VS Code's built-in terminal (Terminal > New Terminal). | Prereq |
| B | Install Paperclip One command |
This single command downloads Paperclip and sets up everything on your computer. It takes about 2 minutes and asks a few yes/no questions. | Paste this into your terminal and press Enter: npx paperclipai onboard --yes |
Install |
| C | Open the dashboard Your control panel |
The dashboard is Paperclip's visual control panel - it opens in your browser like a website, but runs locally on your machine. This is where you'll manage everything from now on. | Open your browser and go to: http://localhost:3100 |
Install |
| D1 | Name your team Create your company |
Paperclip uses a restaurant metaphor: your company is the "restaurant," and your AI agents are the kitchen staff. The name helps you stay organized if you create multiple teams later. | Type any name that makes sense to you. Examples: "Content Team," "Marketing Ops," or "Lu's AI Kitchen." This is just a label - you can change it later. | Config |
| D2 | Write a mission Create your company |
The mission tells your agents what the team exists to do. Vague missions produce vague work. Specific missions produce specific results. | Write one clear sentence. Good: "Produce one LinkedIn post per week about AI tools for consultants." Bad: "Do content stuff." The more specific, the better your agents perform. | Config |
| E1 | Create your first agent Hire the head chef |
Every team needs a leader. Your first agent is the "head chef" - it reads your requests, breaks them into tasks, and delegates work to other agents it hires later. | Click "Create Agent" in the dashboard. Name it something like "CEO" or "Head Chef." This is the only agent you create manually - it will propose and hire the rest. | Config |
| E2 | Pick the AI model Hire the head chef |
The "model" is the brain behind your agent. Smarter models cost more per task. For your head chef, a mid-tier model is the sweet spot - smart enough to coordinate, cheap enough to run daily. | Select Claude Sonnet 4.6 from the dropdown. Best value It's significantly cheaper than the top-tier model (Opus) and handles coordination tasks perfectly. | Config |
| E3 | Test the connection Hire the head chef |
This confirms your API key works and your agent can actually reach Claude's brain. If this fails, nothing else will work - so it's worth checking now. | Set the adapter to "Claude," then click the "Test now" button. You should see a green success message. If it fails, double-check that your API key from step A3 is correct. | Config |
| E4 | Give it a task Hire the head chef |
This is your first real test. You'll see the agent think, plan, and produce output - proof that the whole system works. | Type a simple task in the ticket field, like "Write a 3-sentence summary of what our team does." Click Publish. Watch the agent work in real time. Your first AI employee is live. | Config |
| F1 | Set a budget Safety controls |
Without a budget, agents can burn through API credits with no limit. A budget cap means you'll never wake up to a surprise bill. The agent simply stops when the money runs out. | Go to Settings > Budget. Set your monthly limit to 20€ to start. You can always increase it later once you see how much your agents actually use. Most small teams spend 5-15€/month. | Config |
| F2 | Set a heartbeat Safety controls |
A "heartbeat" is how often your agents wake up and check for work. Without it, agents would run 24/7 - burning tokens even when there's nothing to do. Think of it as office hours for your AI staff. | Set the heartbeat interval to 8 hours. Your agents will wake up, do their tasks, then go to sleep. Three work sessions per day - just like a real team with shifts. | Config |
| F3 | Confirm the safety net Safety controls |
This is the most important safety feature. When an agent hits its budget limit, it doesn't just warn you - it stops completely. No exceptions. No overruns. | No action needed - this is on by default. Just know: when the budget runs out, agents pause automatically. You review what happened, top up if needed, and they resume. You're always in control. | Config |
| G1 | Check your inbox Manage the team |
Your head chef doesn't just execute blindly - it proposes plans first and waits for your approval. The inbox is where those proposals land. You stay the decision-maker. | Open the Inbox tab in your dashboard. You'll see your head chef's plan for the task you assigned. Read it, then click Approve or send it back with notes. Nothing happens without your say-so. | Config |
| G2 | Approve new hires Manage the team |
As tasks get more complex, your head chef will propose hiring specialist agents - a researcher, a writer, an editor. You approve or reject each one, just like a real manager. | When a "hire request" appears in your inbox, review the proposed agent's name, role, and model. Click Approve to add it to the team or Reject if it doesn't make sense. Start small - 3 to 5 agents is plenty. | Config |
| G3 | Watch the org chart Manage the team |
The org chart shows you your whole AI team at a glance - who reports to whom, who's working, who's idle. It updates automatically as agents are hired and tasks flow through the system. | Click the Org Chart tab. You'll see your head chef at the top with any hired agents below. This is your bird's-eye view. You're the restaurant owner now - the kitchen runs itself, and you watch from the pass. | Config |
Want this set up for you instead of by you?
Book a free call →