Problem it solves: an EM gets no structured feedback on how well they run their own 1:1s. This grades the manager's conduct deterministically against best-practice frameworks, tracks development goals over time, and turns "how am I doing on 1:1s" into an evidence-backed scorecard.
Reads a week of already-logged 1:1 sessions plus their transcripts, grades the EM's own conduct (coverage of the 8 dimensions, talk-time, ask-vs-tell, upward-feedback solicitation, prior-item follow-through) with deterministic arithmetic, compares against prior weeks, scores development goals, runs a micro-experiment loop, writes a private Coaching Log row, renders an HTML scorecard, and writes a one-line Next 1:1 Focus onto each report's latest session record. Grades the manager, never the report.
Grades how the EM facilitated a single difficult / conflict conversation or mediation - reads the Fireflies transcript, isolates the EM's own facilitation moves, and scores them deterministically against the 7-dimension conflict rubric with a mandatory verbatim quote per dimension. Grades the manager's facilitation, never who was right.
Grades how the EM facilitated a single agile ceremony or team meeting (standup, retro, planning, review) - reads the transcript, isolates the EM's facilitation moves and non-EM speaker distribution, and scores the 7-dimension facilitation rubric with mandatory verbatim evidence. Grades mechanics, never meeting substance.
Grades the feedback the EM delivered in a single 1:1 - reads the transcript, isolates the EM's own spoken turns, and scores delivery deterministically against the 8-dimension feedback rubric with mandatory verbatim evidence. Grades craft, never the teammate; does not draft the feedback (that's em-feedback-builder).
Grades how the EM ran a single hiring interview - reads the transcript, isolates the EM's own spoken turns (using candidate airtime as evidence), and scores conduct deterministically against the 7-dimension interviewing rubric with mandatory verbatim evidence. Grades the interviewer's craft, never the candidate.
Grades how the EM ran a single live offboarding / exit / role-transition conversation - reads the transcript, isolates the EM's own spoken turns, and scores conduct deterministically against the 7-dimension offboarding rubric with mandatory verbatim evidence. Grades the manager's craft, never the departing person.
Grades how the EM ran a single new-hire onboarding / 30-60-90 kickoff conversation - reads the transcript, isolates the EM's own turns (plus the new hire's where the rubric needs them), and scores facilitation deterministically against the 8-dimension onboarding rubric with mandatory verbatim evidence.
Grades how the EM presented a status update, board/exec presentation, or proposal - reads the transcript including Q&A, isolates the EM's own turns from the audience's, and scores delivery deterministically against the 7-dimension presenting rubric with mandatory verbatim evidence. Grades spoken delivery, never the slides.