OpenFront

OpenFront is an open-source platform for running and coordinating AI agent fleets. I designed the core product surfaces — from daily operations through agent creation and configuration.

The AI SDR is four agents coordinated in sequence. The timeline shows the schedule; the graph shows the dependencies. Same agents, two optical standards — pick the one that answers your question.

The AI SDR timeline and dependency graph — schedule and structure, toggled from the same view.

Type what you want and the system drafts the agent's requirements — inputs, credentials, run frequency. The design challenge: don't skip the review step. Every field surfaced, every assumption visible before it deploys.

ICP Scraping Agent installation — AI drafts the requirements, you review before it deploys.

Every agent runs on a narrative — a goal with steps. The blockers panel shows exactly what's preventing progress and links straight to it. Human intervention only when the agent genuinely can't continue.

AI SDR narrative view — the goal, the steps, and the blockers that need a human.

One agent, multiple jobs. Kits let technical users configure each use case separately — model, toolset, step limit. Configure by conversation or fill in the form. Both paths end in the same place.

Report Agent Kits — per-use-case configuration with chat or form.

AI agents need an interface the way computers needed a desktop. OpenFront is that bet — a central surface where humans direct, monitor, and unblock AI systems. The design problem of the next decade.