llmhyy / microbat

A feedback-based debugger for interactively recommending suspicious step in buggy program execution.
55 stars 16 forks source link

[Project Test] A Benchmark for Testing Agent #150

Open llmhyy opened 6 years ago

llmhyy commented 6 years ago

I find that the regression bugs appear from time to time, from mutation to agent. In the future, we need to set up a benchmark for running result based on d4j projects to see whether new modification introduce bugs on previous results.