turns your codebase into an autoresearch loop — discovers what to measure, instruments the benchmark, then runs tree search with parallel subagents.

a recursive self-improving harness designed to help your agents (and future iterations of those agents) succeed on any task