Michael Richman on FOMAT: Fear of Missing Agent Time

Michael Richman, founder of Cmd+Ctrl, addressed the growing concern of FOMAT, or "Fear of Missing Agent Time," at the AI Engineer Europe conference. He articulated the challenge many developers face: the potential for AI agents to become unavailable or unresponsive precisely when their input or intervention is most critical.

Understanding FOMAT: The Unavailability Anxiety

Richman painted a vivid picture of FOMAT, likening it to the common anxiety of missing out on important events or opportunities. In the context of AI, this fear manifests when an agent, tasked with a complex operation, becomes stuck or requires human input at a time when the user is unavailable. He highlighted that today, agents might require user attention within minutes or hours, but as autonomy increases, this "window" for intervention could expand to days, exacerbating the problem.

The core of the issue, Richman explained, is the unpredictability of AI agent workflows. Users cannot always anticipate when an agent will need their attention, leading to a state of constant, low-level anxiety about being out of sync with their digital assistants. This problem is compounded when users are away from their primary development machines, making it impossible to provide timely input.

The Cmd+Ctrl Solution: Agent Choreography

To combat FOMAT, Richman introduced Cmd+Ctrl, a platform designed to provide centralized management and visibility over AI agent sessions. The system acts as a "control plane" that aggregates all agent sessions, regardless of the underlying framework or agent type. This allows users to monitor the lifecycle and state of their agents, whether they are running on a local machine or in the cloud.

Cmd+Ctrl's architecture involves a "daemon" that monitors agent lifecycles and reports their state. This information is then aggregated by the control plane, providing a unified overview through a user-friendly interface accessible via web, mobile (iOS and Android), or desktop applications.

A Demo of Seamless Session Management

Richman demonstrated the practical application of Cmd+Ctrl, showcasing how users can initiate new sessions, select the agent and workstation, specify a working directory, and define a task. He highlighted the ability to receive notifications and interact with agents remotely, illustrating this with a scenario where an agent, after being given a task on a phone, could be resumed and managed from a desktop terminal or vice-versa. This seamless transition between devices and environments is key to alleviating FOMAT.

The platform's coding-tool agnostic nature was also emphasized, with Cmd+Ctrl supporting a range of popular AI coding tools and frameworks, including Claude Code, Cursor, Codex CLI, Gemini CLI, GitHub Copilot, and OpenCode, with more integrations planned. This flexibility ensures that users can integrate the tool into their existing development ecosystems.

Alleviating FOMAT: The New Flow

Richman contrasted the "old flow" of hyper-focus on a single task with the "new flow" he termed "agent choreography." This new paradigm involves managing multiple agents concurrently, orchestrating their interactions and ensuring they are working effectively at the right times, and crucially, wherever the user happens to be. Cmd+Ctrl aims to provide the necessary tools for this agent choreography, allowing users to be productive without being tethered to a single workstation.

The ultimate goal, Richman concluded, is to reduce the cognitive load associated with managing AI agents, enabling developers to focus on creativity and problem-solving rather than the operational overhead of keeping their agents in sync and responsive. He invited the audience to try the tool and provide feedback, emphasizing the collaborative nature of developing these new AI workflows.