Peer-to-peer (A2A)
Nodes connect by proximity; no centre.
Verified against a2a-protocol.org
A peer-to-peer topology connects agents directly to one another, agent-to-agent (A2A), with no central router; each agent discovers and negotiates with the peers it needs. It is the shape emerging around open agent-interop protocols, where independently-owned agents must collaborate without a shared orchestrator. The upside is autonomy and no single owner; the cost is trust, discovery and the absence of a central place to reason about the whole.
Reasoning cores connect directly over an A2A transport; discovery and negotiation replace a central hub; each agent owns its own perimeter and must authenticate its peers.
- ·Cross-organisation agent collaboration over A2A.
- ·Marketplaces of independently-owned agents.
- ·Federated workflows with no single owner.
- ·No central bottleneck or single owner.
- ·Agents stay autonomous and independently deployable.
- ·Scales across trust boundaries.
- ·Resilient to the loss of any one node.
- ·Trust and authentication between peers is hard.
- ·Discovery and negotiation add overhead.
- ·No central place to audit or bound the whole system.
- ·Emergent global behaviour is hard to predict.
Description, use cases and trade-offs: Field Book analysis, synthesised from established agent-architecture literature.
The corpus does not rewrite vendor documentation. Where copy is quoted verbatim it is marked as such and attributed to its source.