Question 1

What is Hivework?

Accepted Answer

A tool for building, running, and sharing AI agents. Design an agent on a visual canvas, dry-run it, and ship it to a managed runtime. Share a working link with your team.

Question 2

Do I need to know how to code?

Accepted Answer

You shape the agent on a canvas — persona, behavior, tools, output — no framework required. If you already have an agent reachable at a URL, point us at it. If you use a framework like Mastra or LangGraph, you already have what you need.

Question 3

Where does my agent run?

Accepted Answer

On a managed runtime Hivework operates, or on your own endpoint — your call. Bring your own model and keys, or use Hivework-hosted compute when you outgrow your own setup. The build, run, and share surfaces don't change either way.

Question 4

Can I share it with my team?

Accepted Answer

Yes. Ship the agent and send a working link — teammates run it without setup and without a copy of your keys. Self-serve team workspaces are on the near-term roadmap.

Question 5

Can I sell what my agent makes?

Accepted Answer

Yes — that's the next chapter. When you're ready, turn on operator-owned payments: cards via Stripe by default, USDC on Base (x402) for agent-to-agent, and Lightning for pro mode. Hivework never holds the funds. We lead with the build-and-run tool; selling is there when you want it.

Question 6

Is it ready?

Accepted Answer

An early operator is using it in production. The canvas, dry-run, managed dispatch, and webhook delivery work. Team workspaces, the additional payment rails (USDC/x402, Lightning), hosted-compute scale-out, and the public reputation page are still being built. If you're early and curious, build one. If you need polish, give it a month.

Build an agent. Put it to work.

Shape how it works.

Run it.

Ship it.

Put it to work.