reconcile-yesterday

NBA prediction pipeline reconciliation compares actual player participation against cached, featured, and predicted sets to identify mismatches and coverage gaps across all six pipeline phases. It provides actionable remediation steps and backfill orchestration guidance based on the findings.

Backfilling pipeline gaps for NBA predictions involves running a day-after reconciliation that highlights players who played but were not predicted, categorized by HIGH_IMPACT, MEDIUM, and LOW priority. Remediation orchestration then suggests or triggers targeted backfills across pipeline phases to restore data completeness.

Date-specific reconciliation is supported through an override option for a specific date, while the analysis defaults to yesterday's date. You can trigger the reconciliation for a custom date by running the command with the desired date parameter.

NBA data pipeline mismatches between played players and predictions occur when actual player participation does not align with cached, featured, or predicted datasets. The reconciliation process identifies these gaps across all six pipeline phases and provides actionable remediation steps to resolve them.

NBA prediction reconciliation requires no external dependencies or specific environment setup to operate. The skill independently identifies and reports gaps by comparing actual player participation to cached datasets and orchestrating targeted backfills across the pipeline phases.

Identifying HIGH_IMPACT player gaps in an NBA prediction pipeline is limited to yesterday's date by default, with an override option for a specific date. The analysis covers all six pipeline phases but focuses on comparing actual player participation against cached, featured, and predicted sets.