4.5 KiB
description, icon
| description | icon |
|---|---|
| OpenHuman is an orchestrator, not a chatbot: durable agent graphs, visual workflows, sub-agent fleets, a split-brain always-on layer, and end-to-end encrypted agent-to-agent sessions, all in one coherent stack. | sitemap |
The Orchestrator
OpenHuman orchestrating a fleet of agents (its own workers and external harnesses alike) over Signal-encrypted agent-to-agent sessions.
Most harnesses run one agent in one loop. OpenHuman is built as an orchestrator: a stack for coordinating many agents, over long horizons, across machines. It does this durably, observably, and under your control.
Five layers make that real:
1. Graphs, not loops
Every agent turn runs on tinyagents, our open-source graph engine. Multi-step work compiles to state-machine graphs with conditional routing: plan → execute ⇄ review → finalize for delegation, phase DAGs for multi-agent workflow runs, and map-reduce fan-out for parallel workers. All of it has durable checkpointing. A graph can pause mid-run (for your answer, for an approval, for a restart) and resume exactly where it stopped.
2. Sub-agent fleets that don't get lost
The orchestrator spawns specialized sub-agents (up to 3 levels deep), reuses compatible idle workers instead of re-spawning, and routes each to the right model tier: heavy reasoning for the core, a fast burst tier for low-context workers. Reliability is structural: a no-progress circuit breaker stops loops, and stuck children hand back a question (pause + resume on your answer) or an Incomplete root-cause summary, never silence. See the Agent Harness.
3. Workflows you can see
Workflows lift orchestration out of the chat: the agent proposes a typed graph of triggers, agents, tools and conditions; you review it on a canvas and save it. Runs are durable, approval-gated, and fully inspectable step-by-step, powered by open-source tinyflows.
4. An always-on split brain
Inbound traffic hits a fast reflex agent that triages in seconds and hands a deep reasoning core a concise brief; the core does the multi-step work and delegates to workers. The subconscious loop reviews compressed session history and injects steering directives, keeping the always-on layer aligned with your goals, while 20:1 compression keeps week-long sessions bounded.
5. Encrypted orchestration across machines
OpenHuman instances collaborate through tiny.place sessions secured with the Signal protocol: real end-to-end encryption, with keys derived on-device and never persisted. Pairing is consent-based and fails closed: an unlinked agent's message is just a message, never an instruction. Your agent can orchestrate other agents (and be orchestrated) without a server ever seeing plaintext.
Because the transport is plain Signal-encrypted messaging, the other side doesn't have to be OpenHuman. Connect Claude Code, Codex, OpenClaw, Hermes, or anything that can hold a session, and use OpenHuman as the conductor for all of your agents and tools.
What's next: RLMs
The direction we're building toward: Rhai-backed language workflows. These are agents that express orchestration as small programs in a sandboxed REPL, rather than a fixed graph, so control flow itself becomes something the model writes, inspects, and repairs. The graph engine, checkpointing, and trust model above are the substrate for it.
Why this differentiates
| Single-agent harnesses (Claude Code, OpenClaw, Hermes) | OpenHuman | |
|---|---|---|
| Execution model | One loop, one context | Compiled graphs, conditional routing, checkpoint/resume |
| Parallelism | Manual / plugin | Native sub-agent fleets, map-reduce fan-out, worker reuse |
| Automation | Scripts & cron | Visual, durable, approval-gated workflows |
| Always-on | None | Split-brain reflex + reasoning core, subconscious steering |
| Agent-to-agent | None | Signal-encrypted sessions, consent-based pairing, x402 payments |
See also
- Workflows · Subconscious Loop · tiny.place Agent Economy
- Agent Harness: the developer deep-dive on graphs, breakers, journals.
- Agent Coordination tools: the user-facing spawn/delegate surface.