Exploring Replay Vision observations

A scanner is a standing LLM probe over session recordings; each time it runs against a session it records one observation. This skill is about the other half of the loop — reading what the scanners have found and doing something useful with it. For creating or sizing scanners, use [[creating-replay-vision-scanners]].

Mental model

• Scanner → observations. One observation = one scan of one session. There is at most one observation per (scanner, session).
• The finding lives in scanner_result. Its shape depends on the scanner's scanner_type, but it always carries a confidence:
  • monitor → a verdict (yes / no / inconclusive) plus an open-ended observation.
  • classifier → one or more tags from the scanner's label set.
  • scorer → a numeric score on the scanner's scale.
  • summarizer → a free-text summary (optionally with facet embeddings).
• Only succeeded observations carry a finding. Triage the rest by status/error_reason (see below).
• Observations are LLM judgments, not ground truth. One observation is one model's read of one session — corroborate before you act on it.