autoresearch

Automating a research loop with script validation requires a stateful workflow that persists research outputs until explicit validation evidence is present. This Skill runs that stateful loop, using script execution as the gate to ensure only validated outputs are accepted.

A validator-gated research workflow is an automated stateful loop that continues generating artifacts until specific validation criteria are met. It prevents finalization of research outputs unless they pass configured validation modes like script execution or architectural review.

Configuring validation modes for research tasks involves specifying either script validation or prompt and architect validation when initiating the loop. You must define the validation mode and artifact generation settings before execution to ensure outputs meet your specific criteria.

Stateful persistence for research automation is needed when outputs must pass explicit validation criteria before acceptance. It maintains the research loop state continuously, ensuring artifacts are only finalized after successful script execution or architectural review.

Yes, you can use prompt and architect validation instead of scripts for research gating. The automation supports multiple validation modes, allowing you to configure the loop to gate persistence based on architectural review rather than script execution.