Asset scoring engine. Call this before asset selection or allocation. Returns liquidity, structural stability, role classification, and composite quality scores for candidate assets, shaped by the requested portfolio segment.

Market regime classifier. Call this before allocation or rebalancing. Returns volatility, liquidity, sentiment, and allocation authorization inputs for automated portfolio workflows, plus the requested portfolio segment context.

Risk policy engine. Call this when portfolio constraints and segment filters must be computed for a supplied risk tolerance, timeframe, and optional portfolio segment. Returns exposure caps, stablecoin floor, risk budget, authorization constraints, and segment-aware universe guidance.

Portfolio rebalance engine. Call this after allocation or on a monitoring schedule with current holdings. Returns target-vs-current drift and the adjustments required to rebalance within policy constraints for the requested portfolio segment.

Classifies whether a protocol or threat family is active in the SCE corpus. Returns `relevance_level` (NONE | LOW | MEDIUM | HIGH), `relevance_score`, `matched_threat_families`, and `matched_doctrine_tags`. Only call when you already have specific threat context to evaluate — not a default preflight. Pass `threat_families`, `doctrine_tags`, or `protocol_name` to get a meaningful result.

Pre-execution safety check. Returns `mode` (NORMAL | INCIDENT | DEGRADED), `recommended_posture`, `reason_codes`, and per-domain `risk_domains`. If mode !== NORMAL or recommended_posture is require_review or pause, call /agent/x402/sce/risk-evaluate before proceeding. Call this before any protocol interaction, routing, rebalance, or capital movement.

Protocol risk gate. Returns `doctrine_action` (BLOCK | ALLOW | ESCALATE), `risk_level`, `risk_score`, and per-domain `risk_domains`. If doctrine_action === BLOCK, halt the action and return `reason_codes` and `recommended_next` to the user. Recheck after `valid_until`. Call this when continuity-mode returns mode !== NORMAL or recommended_posture is require_review or pause.