autoresearch

Automate AI research projects by using a two-loop architecture that coordinates inner-loop experimentation with outer-loop synthesis. This orchestrates literature review, hypothesis testing, and multi-hypothesis management while routing execution to domain-specific skills.

Structure literature survey and experiment orchestration by initializing a structured workspace, then starting an inner loop with initial hypotheses. The inner loop runs rapid experiment iterations while the outer loop synthesizes results to steer research direction.

Yes, autonomous research experiments require access to domain-specific skills for execution. The orchestration architecture routes experiment tasks to these skills, while also requiring a structured workspace and tooling for progress tracking.

Two-loop architecture works by running rapid experiment iterations in an inner loop with clear optimization targets, while the outer loop synthesizes results and identifies patterns. Continuous operation is supported via heartbeat loops.

Yes, you can generate formal research papers and presentations from autonomous experiment results. The orchestration architecture produces these outputs to communicate research progress and findings after the synthesis loop completes.

Limitations include the strict dependency on external domain-specific skills for execution and the requirement of a structured workspace. The architecture cannot function without proper tooling for progress tracking and formal paper generation.