Trace Logger — temporary diagnostic probe

⚠️ CRITICAL RULE: The log code you insert exists only to collect runtime evidence. It is never committed, never shipped, never kept after analysis. Every single instrumentation statement must be removed before the session ends (or before applying any real fix). Think of it as attaching temporary probes to a circuit — you remove them once you've seen the waveform.

Instrument the call chain with structured trace logs to diagnose runtime behavior.

When to use:

• The bug is not reproducible from a single stack trace (e.g., intermittent, timing-dependent, state-driven)
• A suspected code path must be confirmed or ruled out at runtime
• The order of callbacks, lifecycle events, or async results is unclear
• Static analysis (code reading, CodeGraph exploration) was inconclusive
• You need to observe the actual data flowing through a chain of function calls

When NOT to use:

• The bug has a clear single-frame cause visible in a stack trace (use android-crash-analyzer directly)
• The issue is purely a UI/layout problem (use android-ui-regression-checker)
• The bug is a missing permission or manifest misconfiguration