15 — Automation Reality Scorecard (v_rp_automation_reality_scorecard)

dimension	score
adapter_coverage_score	94
scanner_automation_score	100 (real run proven <26h)
drift_detection_score	100 (detector live)
action_queues_ready	2 classes (EXECUTABLE + HANDLER_READY)
real_actuation_score	0 (REAL_RUN disabled)
human_ai_review_loop_score	100 (handler + preview present)
no_blind_spot_guard	PASS
open_blind_spots	9
p0_p1_blind_spots	2

Verdict: AUTOMATION_REALITY_VERIFIED_WITH_PARTIAL_SOURCES

Computed honestly: no MISSING_ADAPTER (else BLOCKED_BY_ADAPTERS); scanner ran <26h (else PARTIAL_READONLY_HEAVY); but PARTIAL/UNKNOWN sources exist (kb_sop, host_trigger_classes, process_observations, ui_route) → VERIFIED_WITH_PARTIAL_SOURCES rather than a clean VERIFIED.

What this means: the automation is genuinely real (scanner runs daily, adapters fresh, orphan queue auto-rebuilt, drift detected, action handlers exist) — but actuation (REAL_RUN) is off and a handful of sources are declared-partial. The system is read-true and action-ready; it is not yet self-actuating. The only score at 0 is real_actuation, which is authority-gated by design.