Coverage Distribution

Detection ≠ Rewrite ≠ Performance

Detection

Has the diagnostic check been independently reproduced? A check classified as PARTIAL remains PARTIAL even if many tests pass.

Automated Rewrite

Can Precedes automatically transform the code? Currently: 0 verified general rewrites.

Runtime Performance

Does the transformation improve measured performance? This requires independent benchmarking, separate from detection.

These represent different capabilities and must never be conflated.

Checks (… of …)

ID Title Status Languages Evidence