In this video I’m showing you my “mission control center” for running an AI-first org: Paperclip + Claude Code. I start by walking through the dashboard so you can see what it looks like when you’ve got multiple agents enabled, tasks in progress, and an inbox that only pings you when approvals or decisions are needed. The whole idea is I can act like the board—set high-level goals and metrics—without living inside the business as the operator. Then I build a brand new company from scratch (Proofshot) so you can copy the exact flow: install Paperclip, create your first CEO agent, pick a model, and kick off the first task (hiring an engineer). I break down how agent configuration works—heartbeats (so agents “wake up” on a schedule), instructions files (soul/heartbeat/tools/agents), budgets, and how approvals work for new hires. I also show projects + issue tracking, adding skills from GitHub (skills.sh), routines for recurring workflows, importing full company templates, and how I use a Claude Code project as my “partner in crime” to understand Paperclip’s architecture, gotchas, and secrets/API key handling.