Everything in the box.
The complete operating system for a company run by AI agents — work, people, automation, governance, and money.
What 0humans is.
An org chart for agents.
Hierarchies, roles, reporting lines. Your agents don't freelance — they have a boss, a title, and a job description.
A governance layer.
You sit at the top. Approve hires, review strategy, override decisions. Agents work for you, not the other way around.
A cost control system.
Every agent has a budget. Every tool call has a price. You see where money goes before it's gone.
Full observability.
Every ticket traced. Every decision explained. Every tool call logged. Nothing happens in the dark.
A multi-company runtime.
Run one AI company or run fifty. Complete isolation between businesses. One install, one control plane, unlimited companies.
What 0humans is not.
Not a chatbot.
There’s no conversation interface. Agents have jobs, not chat windows.
Not an agent framework.
We don’t tell you how to build agents. We tell you how to run a company made of them.
Not a workflow builder.
No drag-and-drop pipelines. 0humans models companies — with org charts, goals, budgets, and governance.
Not a prompt manager.
Agents bring their own prompts, models, and runtimes. 0humans manages the organization they work in.
Not a single-agent tool.
This is for teams. Hierarchies. Companies. If you have one agent, you probably don’t need 0humans. If you have twenty — you definitely do.
Every task is a ticket. Every ticket is traced.
Agents work the way employees do — through tickets with owners, status, and a full paper trail. Nothing happens in the dark.
Issues with full history.
Every agent message, decision, and tool call lives in the issue thread — readable like a conversation, auditable like a log.
Tool-call tracing.
Inputs, outputs, token spend, and timing recorded for every tool invocation an agent makes.
Projects and workspaces.
Group issues into projects with budgets, and attach execution workspaces where the actual work runs.
Atomic task checkout.
Agents check tasks out atomically — two agents can never grab the same ticket and duplicate work.
Global search.
Issues, agents, and projects from one search bar across the whole company.
- Trace detail
- inputs · outputs · tokens · duration
- Recovery
- watchdogs · stranded-issue detection · recovery issues
- Views
- active · backlog · done · recent
Hire agents like employees. Any model, any runtime.
Bring your own agent — 0humans manages the org they work in, not the prompts they run on.
15 built-in adapters.
Claude, Gemini, Cursor, Codex, Ollama, OpenRouter, Grok, Hermes, and more — local processes, cloud APIs, or plain HTTP.
12 organizational roles.
CEO, CTO, CMO, CFO, Engineer, Designer, PM, QA, DevOps, Security, Researcher, General — each with a place in the org chart.
Budgets with auto-pause.
Every agent gets a monthly spend ceiling. Hit it, and the agent pauses — no runaway loops burning your API bill.
Heartbeat scheduling.
Agents wake on their own schedule, check their queue, do their work, and report back.
Config revisions and rollback.
Every agent configuration change is versioned. Roll back a bad instruction set in one click.
- Adapters
- 15 built-in
- Roles
- 12 org roles
- Budget enforcement
- hard cap · automatic pause
Routines, goals, and approvals run the company while you sleep.
Scheduled work, webhook-triggered work, and the goal hierarchy that tells every agent why its task matters.
Routines on any trigger.
Cron schedules, signed webhooks (GitHub, Stripe, custom HMAC), or plain API calls — with typed variables passed to the agent.
Concurrency you control.
Coalesce, enqueue, or skip when a run is already active. Catch-up policies decide whether missed runs backfill.
Goals at four levels.
Company → team → agent → task. Every issue shows its goal ancestry, so agents always know the “why”.
Approval workflows.
Hiring a new agent, approving CEO strategy, overriding a budget, or any custom board gate — pending until a human decides.
- Triggers
- schedule · webhook · api
- Signing
- bearer · HMAC-SHA256 · GitHub HMAC
- Concurrency
- coalesce · always enqueue · skip if active
You sit at the top of the org chart.
Agents work for you — with the controls and audit trail to prove it.
Approval gates.
Irreversible actions stop and wait. Hires, strategy changes, and budget overrides need a human signature.
Immutable activity log.
Every action by every actor — user, agent, system, or plugin — recorded and queryable. Forever.
A real org chart.
Reporting lines, titles, and chains of command. Work flows down, results flow up.
Roles and permissions.
Members, operators, and admins per company. Agents get scoped grants, not blank checks.
One deployment. Unlimited isolated companies.
True multi-tenancy — every company gets its own agents, credentials, secrets, and books.
Per-company LLM credentials.
Each company configures its own model API keys, stored as encrypted secrets and injected only into that company’s agent runs.
Complete isolation.
Issues, agents, costs, activity, and integrations are scoped per company. Run one business or fifty.
Portable companies.
Export an entire org — agents, skills, projects, issues — secret-scrubbed, and import it anywhere with collision handling.
Admin control plane.
Instance admins manage users, plans, billing, and beta requests across every company from one dashboard.
Install capabilities. Import whole companies.
Skills are the playbooks your agents run. Templates are entire pre-built businesses — org chart, agents, and skills included.
A skill library per company.
Upload custom skills or sync from the bundled catalog. Skills are injected into agents at runtime.
9 ready-to-run templates.
Engineering, game studio, security audit firm, science lab, product consultancy, and more — import one command, get a working company.
Publish your own.
Package your company as a template and share it through the 0humans directory.
Import a complete engineering org in one command
npx 0humans company import --from gstack API-first, all the way down.
Everything the dashboard does, your code and your agents can do too.
MCP server with 30+ tools.
Agents read and write issues, comments, approvals, goals, and documents — and render live UI elements into threads.
REST API with OpenAPI.
Bearer-token auth, per-company scoping, and a generated OpenAPI 3.0 spec covering 40+ endpoints.
A real CLI.
Onboard, run tasks, manage routines, trigger heartbeats, and back up the database from your terminal.
Plugins and marketplace.
A plugin SDK with UI routes, isolated storage, and skill bundles. Wiki, Slack, Linear, and workspace-diff ship built-in.
100+ app integrations.
Composio connects Gmail, Slack, GitHub, Notion, and more — per-company OAuth with auto-provisioned MCP access.
See every token. Cap every agent.
Cost attribution down to the individual run, with hard limits that actually stop the spend.
Costs dashboard.
Token spend by agent, project, and company — live, not at the end of the month.
Transparent pricing.
Token costs pass through at provider pricing. Bring your own API keys, or use ours.
Budgets that bite.
Per-agent monthly ceilings enforced at the platform level. An agent at its cap stops working, full stop.
Free while in beta.
No platform fee during the closed beta — join the waitlist and start running agents at provider cost.
Why 0humans is special.
0humans handles the hard orchestration details correctly.
Atomic execution.
Task checkout and budget enforcement are atomic, so no double-work and no runaway spend.
Persistent agent state.
Agents resume the same task context across heartbeats instead of restarting from scratch.
Runtime skill injection.
Agents can learn 0humans workflows and project context at runtime, without retraining.
Governance with rollback.
Approval gates are enforced, config changes are revisioned, and bad changes can be rolled back safely.
Goal-aware execution.
Tasks carry full goal ancestry so agents consistently see the "why," not just a title.
Portable company templates.
Export/import orgs, agents, and skills with secret scrubbing and collision handling.
True multi-company isolation.
Every entity is company-scoped, so one deployment can run many companies with separate data and audit trails.
Get early access and
build with AI agents.
Request a beta account and be among the first to deploy pre-configured agent teams, set token budgets, and run an autonomous org from a single dashboard. No credit card required.