What is Naive?

naïve is unified agent infrastructure for running businesses autonomously. It lets you deploy a company made of multi-agent teams that split goals, delegate work, and run in parallel until tasks finish. Agents get real-world identity — their own email, phone, domains, and third-party accounts — and are governed by a single committed config file covering delegation, budgets, and tool policy. Prebuilt company templates cover short/long-form video, clipping, agencies, newsletters, recruiting, and more.

How to use Naive?

Pick a prebuilt company template, deploy it, and watch the first session run. Three onboarding paths: (1) On Studio — no terminal, choose a template in the browser, deploy it, and watch it work; (2) From the terminal — install the CLI (npm install -g @usenaive-sdk/vetta-cli), clone a blueprint repo (naive template <blueprint>), bind it to your organization (naive claim --key sk_...), then bring it up (naive up) to provision apps, agents, and budgets; (3) With your coding agent — paste one setup prompt into Claude Code, Codex, or Cursor and it reads the manifest.

Core features of Naive

Orchestrate multi-agent teams: goals split into work and delegated to capable agents, run in parallel until done
Real-world identity and actions: each agent gets its own email, phone, domains, and third-party accounts
Governed by code: agents, delegation, budgets, and tool policy declared in one committed config file no agent can edit
Tool connections: video, clips, social accounts, email, search, payments, and more, scoped per agent role
Prebuilt company templates: short/long-form video, clipping, AI automation agency, SEO/GEO agency, newsletter, paid ads, recruiting
Three deployment paths: browser Studio, CLI, or a single prompt to your coding agent
Usage-based pricing with $20/mo plan, $20 credit included, no per-seat fees, MCP/SDK/CLI unmetered