Agentforce adds Multi-Agent Orchestration
Salesforce documents Multi-Agent Orchestration in Agentforce as generally available for coordinating specialized agents across complex enterprise work.
Salesforce documents Multi-Agent Orchestration in Agentforce as generally available for coordinating specialized agents across complex enterprise work.
Agentforce was documented as an enterprise platform for individual and specialized agents.
Multi-Agent Orchestration is documented as generally available for coordinating specialized agents as a team.
Salesforce introduced a new stage of Agentforce in September 2026. A central change for agent architecture is Multi-Agent Orchestration, which Salesforce describes as generally available for coordinating work across specialized agents when a task spans multiple roles, systems or process stages.
For the AgentenCode profile, this changes the orchestration picture. Rather than equipping only individual agents with tools and business data, Agentforce can coordinate multiple specialized agents as a team for broader enterprise processes.
Salesforce also announced AI Skills for Agentforce Coworker and an Agent Optimizer, with separate pilot or rollout timing. AgentenCode keeps those availability states distinct from Multi-Agent Orchestration, which the provider describes as generally available.
Multi-agent capability is not automatically better for every process. Roles, permissions, responsibilities and escalation paths still need explicit design. Governance also becomes more dependent on traceable handoffs: which agent performed which step, under which identity and with which data.
Primary source
What this update is based on
AgentenCode publishes product changes only after source-first review. The original provider source remains authoritative for current details.
Continue checking
Full context in the Agent Profile.
The profile, trust evidence and history separate documented properties from Unknown.