What problem does it solve? Backend codebases accumulate noisy, mis-leveled, and duplicated log statements that drown out the signals needed to debug production issues, while careless cleanup risks deleting logs that represent important business milestones. ## Core Features & Use Cases - Noise Reduction: Identifies and removes or downgrades logs that narrate function entry, per-row progress, polling loops, and routine successes. - Level Correction: Maps ERROR/WARN/INFO/DEBUG/TRACE semantics onto the project's actual logging library (slog, logrus, zap, winston, pino, and others) without imposing foreign conventions. - Structured Context: Improves event naming and adds correlation identifiers like requestId, traceId, jobId, and durationMs while removing sensitive data such as tokens and PII. - Business Milestone Gate: Pauses to ask for clarification before touching logs that may represent billing, payments, ingestion, or audit milestones. - Use Case: Point it at a service module drowning in per-item INFO logs and duplicate exception logging; it consolidates them into a few structured boundary events while preserving the job_failed error trail. ## Quick Start Ask the assistant to review and improve the logging in the current service module, reducing noise while preserving business milestone logs.
