Agents should be defined like the rest of your infrastructure

Most teams' first AI agents live in a Slack bot, a scheduled script, or a notebook someone forgot to clean up. They work, until they don't — and when something goes wrong, there's no record of what the agent had access to, what it spent, or what it actually did.

NeoSpec started from a simple belief: an agent is a piece of infrastructure, and it deserves to be treated like one. Defined in version control, reviewed like code, deployed automatically, and run with the same discipline you'd expect from any other production system — a budget it can't exceed, permissions scoped to exactly what it needs, and a full record of every action it takes.

That's why an agent in NeoSpec is just a Markdown file with a frontmatter block. No proprietary builder, no vendor lock-in — push it to your repo and it runs, on our managed cloud or entirely on your own infrastructure.

We're building NeoSpec in the open. If that resonates, we'd love to have you try it.