Briefs & runs

Briefs & runs is the archive of everything your agents delivered. Instead of hunting agent by agent, you get one inbox across all of them, with a second tab that lists the runs behind the output.

What this screen is for

Home shows what needs you today. This page is the complete record: every brief you can open, newest first, and every run of the agents visible to you. Use it to catch up after time away, to find last month's brief, or to see what a particular run produced.

The agent filter sits above the list as one row of cards, and those cards never disappear when you deselect them — you can always find your way back to any agent.

What you see

1. Briefs & runs — the page title, with Everything delivered to you, in one place. underneath it. When you have narrowed the App to one project, a scope line names it.
2. Briefs and Runs — the two tabs of the page.
3. The filter bar — a row of cards:
   • All, with the number of agents (4 agents);
   • one card per agent, with how long ago its latest item arrived and, on the Briefs tab, a badge with its unread count;
   • on the Briefs tab, the Unread only toggle at the end of the row.
4. The brief list (Briefs tab) — one card per brief: an unread dot for briefs you have not opened, the title, the first lines of the brief itself, and a line with the agent, how long ago it arrived, the number of findings it covers (6 findings) and its trust tier badge.
5. The run list (Runs tab) — one row per run: how it ended (for example completed or cut by budget), the agent, what started it, how long ago, and how many outputs it published (5 outputs). Runs made as experiments carry an Experiment badge. A row expands in place.
6. Show more — at the foot of the list while there are more records to page in.

How to …

Catch up on unread briefs

1. Open Briefs.
2. Turn on Unread only.
3. Open a brief. Once you have opened it, it no longer counts as unread, here or on its agent's card.

Look at one or several agents

1. Press an agent's card in the filter bar. The list narrows to that agent.
2. Press another agent's card to add it to the selection; press a selected card again to take it out.
3. Press All to clear the selection. Your agent selection carries over when you switch between the Briefs and Runs tabs.

See what a run produced

1. Open Runs.
2. Press a run's row to expand it.
3. Open any of the listed outputs — each carries its type badge, such as Finding or Recommendation. If the run published nothing you can open, the row says This run published nothing you can open. That is a result, not a fault.

How to read it

• The excerpt is the brief's own text. Two briefs for neighbouring weeks can have almost the same title; the first lines tell them apart, and they keep the brief's provenance marks.
• The trust tier badge is the brief's provenance. Governed and Derived say how far you can lean on its numbers — see Trust tiers.
• A run's row starts with how it ended. The same symbols and words are used as on the Agents page, so a run reads the same wherever you meet it.
• "Published nothing you can open" is not an error. A run may find nothing new, or produce output that is not addressed to you.
• The count under Show more is the real total. When the page reaches its limit, the line tells you how many more records exist and suggests the agent filter to reach older ones.

Good to know

• Every row here is a record, not a control. Opening a brief or a run changes nothing except your own read state.
• The archive keeps everything: nothing is ever deleted, and older briefs stay reachable through the agent filter.
• Visibility is exactly your access: you see the briefs and runs of the agents you can reach, and nothing else.
• A brief's link to the run that produced it opens that run's row on the Runs tab, already expanded.

Related

• How to read a brief — what you find when you open a brief
• Home — the short list of what needs you today
• Timeline — the same runs as a time line, with every object under its run
• Trust tiers — what the provenance badge means
• Agents — each agent's health and output at a glance