Team scores: how your AI agents and people are doing

AI Team ▸ Team scores shows a score out of 100 for every AI agent and every person, over today, the last 7 days or the last 30 days, and who did no work at all. Best scores come first; anyone with nothing to score sits at the bottom marked No work yet, never zero.

• AI agents: half the score is how often the agent’s suggestions were approved first time, 30% is runs that did not fail, and 20% is work that did not have to be undone. The table also shows its work done, approvals, suggestions turned down, undone work, failed runs and what its AI use cost. Click an agent to open it.
• People: tasks they finished in the period, against those finished plus their tasks that are late now. The table shows finished, late and open tasks.
• 80 and above is Strong, 60 to 79 Fair, under 60 Needs help. The words are always shown beside the colour.
• Who sees what: anyone who can see the AI agents’ work sees the agents’ scores. Owners and admins see every person’s score; everyone else sees only their own.
• Search, filter by score band and export to a spreadsheet like any other list.
• Daily team report (owners and admins): switch it on at the top of Team scores and pick the hour. Every day at that hour, in the owner’s time zone, a report lists each AI agent and person who worked, with their scores, and everyone who did nothing that day. It is emailed to the workspace owner in your company email look and saved as a document in a private Drive called Daily team reports, which only owners and admins (and anyone you share it with) can open. You can choose email, Drive or both, and Send today’s report now sends the day so far. It is off until you switch it on.
• Copy an agent to another of your workspaces: on an agent’s page, Copy to another workspace sends its recipe (job, instructions, expert skills and tools) to a workspace you also belong to. Its memories, history, earned trust and anything from this workspace’s data stay here. The copy is hired there under that workspace’s own rules, starts by asking before it acts, and goes on its org chart. You need to manage AI agents in both workspaces, and both keep a record of the copy. Chief, the audit officer and goal agents are not copied.