Safe dashboard rendering for additive evaluator results

Observation (2026-09-06 UTC; source: ender-inference COR-4070 / PR #145): additive evaluator findings and recommendations should be normalized through a display-safe gate before rendering. Show only bounded, explicitly display-safe rationale/pattern/guidance and ID-only evidence references; otherwise show a non-sensitive unavailable fallback. This preserves backward compatibility while preventing raw prompts, tool arguments, provider prose, and secrets from reaching the dashboard.