Skip to main content

Module explainability

Module explainability 

Source
Expand description

Layered explanation cards for robot-visible controller decisions.

Cards are intentionally compact by default. Each card has a plain summary plus optional input, evidence, and fallback-contract fields so robot callers can decide whether a decision was expected, degraded, or needs operator action without reverse-engineering scattered metadata fields.

Structs§

ExplanationCard
ExplanationFallbackContract
SearchRobotExplanationInput

Enums§

ExplanationDecision
ExplanationSurface

Constants§

EXPLANATION_CARD_SCHEMA_VERSION

Functions§

cache_admission_card
rebuild_throttle_card
search_fallback_card
search_robot_explanation_cards
semantic_unavailable_card
source_sync_deferral_card