autoresearch

Automate iterative prompt engineering by running binary evaluations, mutating prompts, and selecting changes that increase performance. This autonomous skill improvement loop requires a baseline skill in SKILL.md and defined evaluative criteria to track outcomes.

Autonomous skill improvement is a controlled loop that enhances prompt performance by running binary evals, mutating prompts, and tracking outcomes. It applies to any skill target where evaluative criteria can be defined and experiments can be executed safely.

Set up versioned commits for automated prompt optimization by establishing a controlled loop with a baseline skill in SKILL.md. The system safely mutates prompts and tracks incremental enhancements using version control.

You can use automated prompt mutation on any skill target where evaluative criteria can be defined and experiments can be executed safely. The process requires a baseline skill in SKILL.md to begin the iterative enhancement loop.

To start an autonomous skill improvement loop, you need a baseline skill defined in SKILL.md, a set of evaluative criteria, and a controlled loop with versioned commits to safely track and select performance-increasing changes.

Automated skill evaluation loops require versioned commits to safely track outcomes and select prompt mutations that increase performance. This controlled environment ensures repeatable optimization and accurate baseline tracking across iterative enhancements.