Team alignment health check, run for STUD.com (the dogfood workspace).

Pillars, in intake order: Joint Objectives, Joint Commitments, Joint Resources, Joint Risks

(the Team Alignment Map). Each member rates each pillar on the intake slider

(scale_min to scale_max, default -2 to 2). The score is the per-pillar team mean

scaled x100; divergence is the population standard deviation x100 (how split the

team is); below_neutral flags pillars whose mean is under 0 (exactly 0 is not a

gap); weakest is the lowest-mean pillar, ties resolved to the earliest pillar.

Empty or ragged rating matrices return the invalid sentinel instead of a score.

def score_alignment(inp): votes = inp.get("votes") if not isinstance(votes, list) or len(votes) == 0: return {"error": "invalid_matrix"} width = len(votes[0]) if isinstance(votes[0], list) else -1 if width <= 0 or any((not isinstance(row, list)) or len(row) != width for row in votes): return {"error": "invalid_matrix"} n = len(votes) means, divergence = [], [] for p in range(width): col = [float(votes[m][p]) for m in range(n)] mu = sum(col) / n var = sum((x - mu) ** 2 for x in col) / n means.append(round(mu * 100)) divergence.append(round((var ** 0.5) * 100)) below = [p for p in range(width) if means[p] < 0] weakest = min(range(width), key=lambda p: (means[p], p)) return {"means": means, "below_neutral": below, "divergence": divergence, "weakest": weakest}