Welcome to Lolabots

The AI agent revolution is here, and most of it is being built for the wrong people.

Walk into any AI agent demo today and you’ll see CLIs, MCPs, API keys, Docker containers, config files, prompt engineering, and the unspoken assumption that you can debug a stack trace if something breaks. That tax buys you incredible power. It also locks out anyone who isn’t already an engineer.

We think the people who would benefit most from a tireless assistant are exactly the people who can’t pay that tax. The small business owner who runs five inboxes. The solo creator who’d kill for an editor that never sleeps. The household manager juggling appointments, school forms, and family logistics. The operator who already has too many tabs open.

So we’re building lolabots.

What lolabots is

Lolabots is a family of practical AI tools. The first one is Lolabot Factory — pick a personality, equip it with skills, paste one command, your agent goes to work. 314 specialists across 18 divisions, skills from Anthropic, Vercel, Google, Cloudflare, gstack, and more. Works in Claude Code, Cursor, ChatGPT, Claude Projects, Custom GPTs, OpenAI Assistants, or any tool that takes a system prompt.

More products are coming. Different problems, same thesis.

What we believe

• Install is one paste. If we need a second command to make it work, we did it wrong.
• The agent handles its own homework. If it needs a CLI to do a job, it installs the CLI when asked. We don’t leave that to you.
• Open beats proprietary. Every skill we ship is source-visible. Our forks live in 23blocks-OS.
• Useful beats clever. No demos that don’t survive contact with reality.

What’s next

This blog is where we’ll announce new lolabots, share design notes, and occasionally argue why AI tooling should look more like consumer software and less like Kubernetes.

Subscribe with your RSS reader of choice, or just bookmark lolabots.com/blog. We’ll write when we have something to say.

— The Lolabots Team