Methodology · v2.8

BCQQualité du conseil

Board composition quality assessment for {legal_name}. EXTRACTION RULES: return type="composite_signal" with one entry per dimension, each carrying a 0-100 LLM assessment + a source_url backing that specific dimension. Required sub-dimensions (use these exact names): (a) independent_director_pct — 0-100 = percentage of independent directors out of total board (cite proxy / DEF 14A / governance report); (b) diversity_score — 0-100 composite of gender + nationality + skills-matrix diversity (cite proxy + ISS / Glass Lewis / MSCI Governance / BoardEx); (c) avg_meeting_attendance_pct — 0-100 = average disclosed board-meeting attendance %; (d) skills_matrix_completeness — 0-100 = how granular and disclosed the skills matrix is. Each signals[].value MUST be a real assessment based on cited evidence (NOT 0 placeholder). Server applies composite_signal_score (weighted-average of sub-signals).