What it is
Agent mesh is an architectural pattern where independent agents (sales, service, HR, third-party) discover, communicate, and hand off work to each other via shared protocols. It extends microservices thinking to autonomous agents at enterprise scale.
Why it matters
Mid-market companies rarely run one agent — they run several. A mesh prevents siloed agents from duplicating work and enables end-to-end automation across business functions.
Key components
- agent discovery
- handoff protocol
- capability registry
- cross-agent orchestration
How it connects
Agentforce agents can participate in a mesh using protocols like A2A and MCP, registering capabilities so other agents can invoke them safely.
Good to know
Start small: a mesh only pays off once you have 3+ agents that genuinely need to collaborate, not before.
Related terms
Agent Handoff Protocol
The structured rules governing how one AI agent transfers context and control to another agent or human.
Agent Orchestration
The coordination and management of multiple AI agents working together to accomplish complex workflows that no single agent could handle alone.
Capability Registry
A structured catalog that maps AI capabilities (reasoning, structured output, tool use, vision, long context) to the models that can serve them — the substrate that makes skills portable across LLM vendors.
Multi-Agent System
An architecture where multiple specialized AI agents collaborate to handle complex workflows — each agent owns a domain and they hand off work to each other.
