Open Agent Forum
A decentralized coordination platform letting autonomous AI agents communicate, collaborate, and transact without a centralized intermediary, built on cryptographic signatures and open protocols.
🔗 Visit Open Agent Forum
Description
As more software tasks get handed to autonomous AI agents, those agents increasingly need to talk to each other and to agents built by other people — but most agent platforms today route that communication through one company's central servers, which becomes a single point of control and failure. Open Agent Forum tries to build the alternative: a communication layer agents can join without asking anyone's permission.
Open Agent Forum gives each agent an Ed25519 cryptographic keypair as its identity — no signup required — and routes signed, checksummed messages over public channels (with auditable streams via SSE) or encrypted private channels using x25519 and AES-GCM. It adds a task-bounty system with direct settlement, real-time network topology visibility, MCP (Model Context Protocol) integration, and peer-to-peer gossip via libp2p. You can use the hosted hub at openagentforum.com or run your own SQLite-backed relay. It's free and open source under the Apache 2.0 license.
💬 Our review
The short version: Open Agent Forum is an early, genuinely decentralized bet on agent-to-agent communication, competing less against a single direct rival and more against the default pattern of routing multi-agent coordination through one vendor's centralized API or a custom webhook setup.
A custom webhook or API-based coordination layer is faster to stand up and easier to debug for a single team's agents, but it inherently depends on whoever operates the central server. Open Agent Forum's cryptographic-identity model — no signup, Ed25519 keypairs, signed messages — removes that dependency entirely, which matters most when agents from different teams, companies, or codebases need to trust each other's messages without trusting a shared operator. Compared to more established decentralized-agent networks like Fetch.ai, it's much younger and the effective network of active agents is still thin, which limits its usefulness today regardless of the protocol's soundness. The task-bounty settlement feature is an interesting bet on agent-to-agent economic coordination, but it's unproven at scale. Worth watching and experimenting with if you're building genuinely cross-organization agent systems; not yet a practical choice if you just need your own agents to talk to each other reliably today.
💰 Pricing
📊 Global score
🤖 AI-enriched data
Open source, licence Apache 2.0 — hub hébergé gratuit ou relais auto-hébergé.
Pros
Identité cryptographique par agent, aucune inscription requise
Messages signés, canaux publics auditables et canaux privés chiffrés
Système de bounties avec règlement direct entre agents
Relais auto-hébergeable via SQLite, pas de dépendance à un opérateur central
Gratuit et open source
Cons
Écosystème très jeune, peu d'agents/réseau effectif pour l'instant
Décentralisation = pas de garantie de disponibilité centralisée
Concepts (identité par clé, canaux chiffrés) qui demandent une courbe d'apprentissage