Global Truth Safety

Verify all assertions about code behavior by linking them to test output, logs, or benchmarks. Document which tests validate each claim and reference specific log entries or metrics that confirm the code works as stated, preventing unverified assumptions from reaching production.

Flag issues the moment you discover them rather than deferring or hiding them. Document the blocker with concrete evidence—error messages, test failures, or system logs—so the team can address it promptly and prevent cascading failures.

Report only facts you can validate: tested features, observed system behavior, and documented gaps. Cite logs, test results, and specifications rather than assumptions, creating an auditable trail of what actually works versus what remains untested or at risk.

Yes. Before shipping a feature, confirm all behavior claims with passing tests and observed system logs. Challenge any assumption lacking evidence, document test coverage, and flag gaps so stakeholders know exactly what has been verified and what remains uncertain.

Unverified claims hide bugs, waste rework, and erode trust. Data-backed reporting catches regressions early, lets teams make informed decisions, and builds a culture where status updates and code assertions are trustworthy because they rest on logs, benchmarks, and test evidence.