1. Prepare
Use representative, complete evidence and include the context needed to distinguish a real issue from an expected variation.
Review CPI retry evidence with conservative safeguards. The structured settings go to OctaSapien for analysis; no SAP call or replay is made.
Practical usage guide
Assess structured CPI, JMS, MPL, receiver and idempotency evidence for a conservative human retry decision. No SAP calls or replay execution.
Use representative, complete evidence and include the context needed to distinguish a real issue from an expected variation.
Prioritize concrete findings, corroborate them against the source and rerun the analysis after correcting the input.
Validate the result against the interface contract and target CPI runtime before transporting or sharing it.