FlowSharp supports the Model Context Protocol (MCP), the emerging standard for connecting AI agents to business systems and tools.
What FlowSharp's MCP interface enables:
Process querying — an AI agent connected via MCP can ask "How many cases are stuck in 'Awaiting data'?" or "What is the average time in the review phase?" and receive a structured response from real data.
Flow triggering — the agent can start a new item in a FlowSharp flow on user request, for example "Open a new onboarding case for client ACME".
Task completion — an AI agent can receive a task assigned via MCP and complete it with a structured output, which FlowSharp validates and records in the timeline.
Item unblocking — Admins can unblock stuck items directly from the AI agent interface, without accessing the FlowSharp console.
Security:
- MCP access respects the FlowSharp roles of the connected agent
- Data Shield applies to MCP queries too: sensitive fields are never exposed
- Every MCP action is audited in the timeline with agent identification
The MCP integration is designed for the Flow Agent: the integrated assistant that helps operators and managers query and act on processes in natural language, maintaining governance and traceability.