autoresearch-create

Autonomous optimization experiments are configured by defining goals, metrics, scope files, and constraints. This skill generates the necessary configuration files, manages git workflows, and initiates iterative testing across project files to continuously improve code or process metrics.

Automating iterative testing and benchmarking involves creating an experiment loop that tests modifications and logs progress automatically. This skill establishes the loop, takes baseline measurements, and benchmarks results iteratively to reduce manual setup time and improve the build process.

Git workflow integration is fully supported for autonomous experiment scripting. The setup process automatically manages git workflows alongside the creation of experiment scripts and configuration files, ensuring version control is maintained throughout the optimization iterations.

Defining user-defined constraints is required before running autonomous optimization loops to ensure safety. The session configuration guides you through establishing goals, metrics, scope files, and boundaries that the automated testing and modification process must respect.

Starting an automated experiment loop requires configuration files like autoresearch.md and autoresearch.sh. These files are generated automatically during setup to document the experiment state, execute the optimization scripts, and manage baseline measurements.

Autonomous experiment loops should not be used when you cannot establish clear user-defined constraints or safety boundaries. Without defined metrics, scope files, and specific optimization targets, the automated modifications and iterative testing cannot be safely controlled.