Self-benchmark: YOU write the code, adversarial reviews it (multi-provider), you fix, you write tests, adversarial reviews tests, you fix. Measures YOUR quality as an agent. Run in different models (Opus, Sonnet, Haiku) and compare results.
Self-benchmark: YOU write the code, adversarial reviews it (multi-provider), you fix, you write tests, adversarial reviews tests, you fix. Measures YOUR quality as an agent. Run in different models (Opus, Sonnet, Haiku) and compare results.
Esta entrada no publica ningún paquete de npm, así que Forge no tiene un árbol de dependencias para ella. Es una carencia de cobertura, no una afirmación de que no tenga dependencias.