Federated organization
User nodes, agent nodes, and service nodes can run on different machines while staying inside one governed graph.
Federated AI organization
Run a federated AI organization.
Entangle is not another agent orchestrator. It is a runtime for distributed organizations of agents and users: signed nodes on different machines, coordinated over Nostr, sharing work through git, governed from one graph.
Relay-based coordination
Signed Nostr messages let nodes coordinate without requiring public inbound network access for every runner.
Git-backed work
Artifacts are durable refs: code changes, reports, source updates, and knowledge can be published, reviewed, and restored.
Governed placement
Nodes are assigned to trusted runners, and the live operating view is built from signed runtime observations.
Federated organization graph
Agents and users as nodes
Nostr + git substrate
Observable by design
Why Entangle
Agent orchestration is not enough.
Most systems put a central orchestrator in front of agents. Entangle treats the organization itself as distributed infrastructure: agents and users can live anywhere, sign their own messages, and exchange durable work without collapsing into one hidden controller.
Organization
Agents and users together
Humans are not outside the system. They are graph nodes with identity, routes, approvals, and conversations.
Distribution
Runners anywhere
Agent and human-interface runtimes can be assigned to runners on different machines and networks.
Federation
No public IP dependency
Nodes coordinate through relays and signed events, so runners do not need to be directly reachable.
Substrate
Git carries the work
Agents hand off artifacts, source changes, and knowledge through git refs instead of message blobs.
Runtime primitives
Six primitives for federated AI work.
Entangle keeps topology, identity, execution, coordination, artifacts, and observability separate enough to distribute and connected enough to govern.
Users are nodes
Human participants have stable identities, sign their own messages, and live inside the same distributed graph as the agents.
Edges govern work
Typed relationships decide who can delegate, review, approve, escalate, or hand off. The org chart is executable.
Nostr connects the organization
Tasks, replies, approvals, and lifecycle events move as signed Nostr messages between nodes that can live on different machines.
Artifacts move through git
Work products are committed, published, fetched, inspected, and restored through real git references.
Runners can live anywhere
A runner joins with a signed handshake, advertises capabilities, receives assignments, and runs nodes without public IP gymnastics.
Operations stay visible
Nodes, runners, sessions, turns, approvals, artifacts, memory, and recovery all surface in Studio and CLI.
Operate it
Boot the graph and watch work move.
Start the runtime, open Studio, inspect the graph, send a task from a user node, watch an assigned agent run, publish a git artifact, and review the trace from the same control room.
1. Open Studio
See the graph, Host status, runner placement, assignments, sessions, approvals, artifacts, and runtime health.
2. Assign and run
The Host assigns nodes to capable runners. The relay carries signed messages. Each runtime proves what it did.
3. Review the artifact
The result lands as a git-backed artifact with lineage, retrieval state, approval context, and trace evidence.
Notes
Field notes from the runtime.
RuntimeOperations
Run the graph
The fastest way to understand Entangle is to boot the runtime and watch work move through the graph.
Apr 26, 2026
VisionRuntime
The team is the runtime
Entangle turns distributed agents, users, services, runners, messages, approvals, and artifacts into one governed federation.
Apr 26, 2026
ArchitectureFederation
Why federation, not orchestration
Agent orchestration puts every actor behind one controller. Entangle is a federated runtime for distributed agents and users.
Apr 26, 2026
Run a federated AI organization.
Boot Entangle, open Studio, and watch distributed agents, users, runners, signed messages, approvals, and git-backed artifacts move through one governed graph.