The idea is simple: before starting work, the agent creates a run, returns a public link to the user, and then posts short progress updates as it works. The user can watch a live timeline instead of waiting in the dark for a long task to finish.
The product is intentionally minimal. There are no accounts, no setup flow, and no dashboard. The homepage just gives you a prompt to paste into your agent. The agent reads the instructions, creates a run, and starts updating it. The public page shows the title, status, timestamps, and event timeline.
The main use case is making agent workflows legible to users. If an agent is researching vendors, debugging a system, reviewing documents, or drafting something large, it can expose a live status page without building a whole product around “agent transparency.”
It was also useful to dogfood while building. I used the product itself to track debugging and deployment work on the app, which helped clarify what kinds of updates are actually useful versus noisy.
Would be interested in feedback on the product framing and whether this should stay extremely narrow or grow into something more general for agent task visibility.