This connects Claude to Agentwork's AI agent delegation platform, letting you spin up and monitor AI agents that can work on tasks while keeping a human in the loop. You can create agents, assign them work, check their progress, and review their output before finalizing anything. Reach for this when you need to farm out repetitive or time-consuming tasks to AI workers but want oversight and approval gates built in. The human-in-the-loop pattern means you stay in control while the agents do the grunt work. Available over both stdio and streamable HTTP transports.
claude mcp add --transport stdio io.github.ferm-ai-agentwork-mcp uvx agentwork-mcp