Comment by rickette

Interesting but how is an "autoresearch loop" different than creating a spec and X number of testcases and letting an agent run against these testcases and the spec?