Team Inspection
The reps on the teams you own, ranked by attention risk, each with a coaching action. Metadata only, never rep content.
Included with leadercron . A seat without it sees an upsell here instead of the page.
Overview
Team Inspection shows a manager which rep needs a 1:1 first. It ranks the reps on the teams you own by attention risk and puts one coaching action on each row, so the manager's scarce time goes where it matters most.
Metadata only, by design
This board reads run and outcome aggregates only, run counts, error counts, last run time, and role coverage, and never a rep's drafts or message text. The team boundary is resolved from teams you own, so no one can widen it to another org. The score maps those aggregates into an attention signal: a rep who has gone quiet in the tools, whose run volume is low, whose runs keep erroring, or who works only one product surfaces first. The privacy boundary is a security property, not a setting.
What each row shows
Every rep row carries the aggregates the score was built from: run count, last run time, distinct roles used, error count, the attention score with its band, and the coaching action from the top driver. The rep's email shows for identity so you know who to book the 1:1 with, that is who is on your team, not what they wrote. Four summary tiles sit above the table: team size, how many reps are above the low-risk band, the average attention score, and the team's open pipeline (a metadata-safe sum of numbers, shown as an em dash until your reps have pipeline-summary rows).
Where the activity comes from
The board reads the run telemetry your reps' scheduled agents already produce. Add teammates in Team admin and their run activity appears here, ranked by attention. A leader who owns no team sees an empty board.
What is not shown
The design mockup sketched a fuller manager cockpit: a deal-by-deal risk table with account names and amounts, per-rep MEDDIC/SPICED scores, a forecast-governance strip comparing each rep's commit and best-case to a model, and a win-loss pattern card. Those are omitted here on purpose. Every one of them is rep deal content, and this seat is metadata-only by design, so rendering them would mean reading, or inventing, data the boundary forbids. The attention ranking, built from run and outcome aggregates, is the honest signal this seat can carry.