differential-fuzzer

You can validate SQL query results by running cross-engine differential fuzzing, which automatically generates SQL statements and compares Turso outputs against SQLite to flag result mismatches.

Cross-engine differential fuzzing is a validation technique that generates broad sets of SQL statements and compares results between Turso and SQLite to catch correctness bugs early in development.

You can run continuous loop fuzzing by executing the differential_fuzzer binary via cargo run, using the --seed flag to adjust generation and -n to control the number of SQL statements executed.

Yes, you can reproduce failures by using the --keep-files flag, which preserves output artifacts including the failed SQL statements and database schema at the point of failure for debugging.

Yes, the Rust toolchain is required to run the differential_fuzzer binary locally or in CI environments, as the fuzzing workflow is built and executed through cargo.

Differences occur when Turso outputs mismatch SQLite results during automated SQL generation, indicating a correctness bug that the differential fuzzer records for reproduction and debugging.