Safe evaluator result APIs require a second public projection

A durable redacted evaluator record should still be treated as untrusted when exposed through a results API. Rebuild the public payload from controlled enums, server-owned rationales, bounded typed references, and allowlisted version identifiers; do not pass through prompts, transcripts, tool arguments, provider prose, or arbitrary provenance. For cross-call remediation, aggregate only exact compatible node/version strata and suppress suggestions when node identity or prompt/schema versions are unresolved. Recompute summary scores from the projected evidence rather than trusting persisted summary fields, so public counts, labels, and evidence cannot contradict each other.