Bioinfoysis is a multi‑agent harness that treats each request as a persistent, artifact‑grounded analysis run. It blends global planning with step‑wise, evidence‑driven replanning: the planner keeps an executable checklist and revises pending steps using structured handoffs returned after each worker execution. These handoffs bind intermediate results to the responsible agent, checklist step, and plan generation, preventing stale evidence from being silently reused after replanning. A controlled runtime validates generated scripts, tables, and figures before they are used downstream; role‑specific context, persistent memory, and governed bioinformatics skills enable reliable execution over long analysis trajectories. We evaluated Bioinfoysis on BixBench and two QA tracks of LAB‑Bench 2. Bioinfoysis achieved a state‑of‑the‑art 82.4% accuracy on BixBench; on SeqQA2 the average accuracy across four language models rose from 27.81% to 64.13%, and on DbQA2 from 3.13% to 31.25%. These results show that reliable bioinformatics automation depends not only on model capability but also on the harness that governs planning, execution, memory, and evidence flow. We hope Bioinfoysis will drive progress in the bioinformatics community. Demo site: https://report.bioinfoysis.com/.
Review