Improve the models and prompting strategies behind automated pull request review. Own evaluation: decide what a good review comment actually is, and measure it.

Requirements

• —Experience evaluating LLM output quality
• —Familiarity with static analysis or code review tooling
• —Python and TypeScript