CHANGELOG

What Shipped This Week: Duty Run History

Your Research agent finishes its 7am competitor scan while you are still asleep. Before this week, you trusted the notification — or opened ~/files and hoped the CSV updated. Now you open the Agents app, expand the duty, and read every past run: status, duration, output snippet, errors if something broke.

5–7 min read

We shipped duty run history and immediate feedback when you manually start a duty. New accounts get a signup-to-first-message path through Cloudia. Mobile got an Agents tab rename, a one-tap hire button, and thread-drawer fixes that were blocking chat on phones.

Every duty run now leaves a paper trail on the Team board

A seven-person growth agency runs a weekday competitor pricing duty at 6:45am. The agent writes a summary and drops a CSV in ~/files/reports/. When something goes wrong — a site blocks the browser, a selector breaks, the duty times out — the founder used to find out from a missing file or a vague push notification. There was no in-app log of what the duty actually did on Tuesday versus Wednesday.

That changed this week. Open the Agents app, pick a specialist, expand any duty on the Team board, and you will see a run history panel. Each entry shows the status (success or error), when it ran, how long it took, a trimmed output snippet, and the error message if the run failed. No more guessing whether the 7am duty actually finished or silently died.

Manual runs got the same attention. Tap Run Now on a duty and the button immediately shows a Starting… state while we poll the server until the run is live. Before, you clicked Run and stared at the board wondering if anything happened. Small interaction, big trust difference when you are testing a new duty before scheduling it for real.

This matters because CloudAxis is built around recurring duties, not one-off chat prompts. Agents accumulate files across runs; pipelines hand off via paths like ~/files/reports/competitor-prices.csv. If you cannot see whether yesterday's duty succeeded, you cannot trust tomorrow's handoff. History closes that loop inside the OS — the same place you configure schedules and hire specialists. We wrote about the always-on model in our guide to scheduling AI work that runs while you sleep; this week the Team board finally shows you what actually ran overnight.

On mobile, the history panel scrolls inside the duty card. Status chips use the same green/red styling as desktop. If a run produced a long summary, we show the first 2,000 characters — enough to spot "23 prices tracked" versus "login page blocked" without loading a full thread. Tap through to the chat or Files app if you need the full artifact.

Next for this area: tighter links from a finished-duty notification straight to the run log and the output file path. For now, history lives where you already manage duties — no new app, no export step.

Also shipped

What's next

We are wiring the duty-finished notification to open run history and the output file in one tap — the Specialist → Duty → Output loop from last month's roadmap. If you are setting up your first recurring duty, the 24/7 scheduling guide still walks through cron, handoffs, and file paths.

Try it at app.cloudaxis.ai