What problem does it solve? Answering "where is the observability loop?" normally requires querying live backends or re-running a stack. This Skill computes the loop's state entirely offline from the repository's committed .odd/ report history and git metadata, producing a one-screen status or full working tables without any backend queries. ## Core Features & Use Cases - Offline loop status: Renders per-lineage loop state, findings burn-down (open, fixed-and-verified, regressed, declined), trends, and open telemetry gaps from stored observation and instrumentation reports. - Verdict and next action: Computes an ok/warning/error verdict with reasons and the next recommended action per lineage, citing evidence for every row. - Judgment workflow: Defers ambiguous rulings to a "Judgment needed" list, then applies caller rulings via flags (--ruled, --runtime, --repository) and re-renders with one truth. - Use Case: After several observation and verification runs, ask where the loop stands for a service on a given stack and environment, and get the burn-down, trends, and the next command to run without touching any live system. ## Quick Start Ask the assistant to show the current ODD loop status for this repository, optionally scoped to a service, stack, or environment.