Paperclip AI is an open-source platform for orchestrating and managing teams of AI agents. Rather than being an AI model itself, it acts as a control center that lets you organize multiple AI agents as if they were employees in a company.
How it works
Instead of asking a single chatbot to do everything, you create an organization such as:
-
CEO
-
CTO
-
Software Engineers
-
QA Engineers
-
Technical Writer
-
Marketing Manager
-
Customer Support
Each AI agent has:
-
A specific role
-
Goals and responsibilities
-
Reporting relationships
-
Budgets
-
Scheduled "heartbeats" where they wake up, perform work, and go idle again.
Features
-
✅ Organize AI agents into an org chart
-
✅ Delegate work automatically
-
✅ Track tasks and progress
-
✅ Audit every decision
-
✅ Set spending limits
-
✅ Run multiple AI-powered companies from one server
-
✅ Use different AI providers simultaneously (Claude Code, Codex, Cursor, OpenClaw, shell scripts, HTTP services, etc.)
Example software company
You might give the CEO a goal like:
"Build a Bootstrap 5 bill management application with MariaDB."
The CEO creates tasks for the CTO, who assigns work to developers. Developers generate code, QA tests it, and documentation is produced automatically. You review and approve major decisions before deployment.
Is it an LLM?
No.
Paperclip does not include its own language model. Instead, it manages other AI systems, such as:
-
Claude Code
-
OpenAI Codex
-
Cursor
-
OpenClaw
-
Local CLI agents
-
Custom scripts or HTTP-based agents
Self-hosting
Paperclip is designed to be self-hosted. It provides:
-
A web dashboard
-
A Node.js backend
-
A React frontend
-
Database support
-
Authentication
-
Agent scheduling
-
Logging and auditing
Would it fit your workflow?
Based on your recent projects—Bootstrap 5 applications, MariaDB databases, PHP/Smarty development, and local AI on Gentoo—Paperclip could be useful if you want to coordinate multiple coding agents on a single project instead of prompting one model at a time. For example, you could dedicate separate agents to:
-
Database schema design
-
Bootstrap 5 UI development
-
PHP backend coding
-
C++ utilities
-
Documentation
-
Testing
This can help parallelize work while keeping tasks organized under one dashboard.
If you're running local models on your Gentoo system with llama.cpp or Ollama, I can also show you how to integrate those models with Paperclip so your AI team runs primarily on your own hardware rather than relying entirely on cloud APIs.