autoresearch-create

Automating an optimization loop requires configuring a session with target metrics, executing commands, and logging outcomes. This skill automates autonomous experimentation loops to continuously test ideas and retain the best results for cyclic optimization tasks.

Autonomous experimentation is a cyclic process of testing variations to improve a chosen baseline metric over time. It works by executing commands, measuring outcomes, and retaining improvements while regressing unsuccessful attempts using persistent guardrails.

You configure experiments by defining the session name, metric, unit, and optimization direction. Run experiments by executing a command, capturing output for evaluation, and recording results with a metrics dictionary to support auto-commit and rollback.

Yes, this handles cyclic optimization tasks across data science, product experiments, and software optimization. It is applicable wherever a baseline can be improved over time through automated testing and persistent loop execution.

You can resume an interrupted automation loop using context-resume guidance. The loop operates with guardrails for improvement versus regression and maintains associated documentation for audit trails and easy rollback.

Autonomous optimization loops require a clear baseline and measurable target metric to function. You must reinitialize the experiment session whenever the optimization target changes, and rely on auto-commit and rollback for error handling.