omc-autoresearch

Automate iterative research loops by defining user objectives, success criteria, and an evaluator to guide the progression. The tool manages structured data evaluations and durable workspace artifacts to streamline continuous improvement processes.

Starting an evaluator-driven improvement loop requires defining specific user objectives, success criteria, and an evaluator. You must also specify a mission directory and maximum runtime duration to bound the automated research process.

Structured data evaluation guides continuous improvement by using defined success criteria and an evaluator to assess each iteration. This feedback directs the loop's progression, ensuring software development features meet required objectives.

Yes, you can manage research artifacts locally. The automated iterative research process utilizes workspace-local artifacts for tracking and referencing research activities throughout the continuous improvement loop.

Bounded research loops require a predefined maximum runtime and rely entirely on the user-defined evaluator's quality. The automated process cannot progress without explicit success criteria and structured data evaluations to guide it.