systematic-debugging

Systematic debugging forces an evidence-first workflow that captures exact reproduction commands, full stdout/stderr, and environment facts to localize test failures before proposing any fix. You must reproduce the bug, narrow the root cause, and verify the correction with regression evidence.

Fix flaky behavior and runtime errors by capturing complete error output, exit codes, and timestamps, then applying an evidence-based hypothesis test to isolate the root cause. Once narrowed, implement the minimal fix and add a regression guard to verify the correction holds across boundary conditions.

Reproduce integration boundary failures by replaying exact commands and logging full environment facts and deltas to isolate platform-specific behaviors. Systematic debugging gates progress until reproduction evidence and affected contracts are fully captured and accessible.

Stop the line during root-cause analysis when reproduction evidence, required access, affected contracts, or regression guard inputs are missing. This safety mechanism prevents guesswork by blocking progress until boundary evidence is fully documented and accessible.

Ensure minimal fixes for runtime issues by applying the smallest change that addresses the narrowed root cause, then verifying with the original reproduction command and adding an appropriate regression guard or documentation handoff to prevent future regressions.