The results suggest that integrating persistent execution with session-level provenance can improve the reliability of AI-assisted scientific workflows, and recommend that integrating persistent execution with session-level provenance can improve the reliability of AI-assisted scientific workflows.
Abstract
AI co-scientists could accelerate computational research, but over a long-running study the workflow also has to stay inspectable, resumable and reproducible, which requires persistent computational state and provenance. Here we present OpenAI4S, an open-source scientific research agent built around the principle of \emph{Code as Action, Science as Sessions}. OpenAI4S combines a persistent computing runtime with research-session management: orchestration is handled through structured tool calls, while scientific actions are represented as complete code cells executed in persistent Python and R kernels. An append-only Action Ledger, per-cell execution records, versioned artifacts, environment records, and workspace checkpoints preserve how results were produced and support session recovery, branching, and extension. Configurable sandboxing, permission controls, and code and trajectory screening provide complementary safeguards. We evaluate OpenAI4S on 36 research scenarios spanning retrosynthesis, molecular dynamics, protein binder design, protein mutation, catalyst screening, and mineral spectroscopy, measuring scientific task accuracy, workflow completeness, and reproducibility of the resulting repositories. OpenAI4S achieves an overall score of 7.83, compared with 5.7--6.4 for a general-purpose coding harness evaluated with three frontier models, with the largest gains on long-horizon and computation-intensive workflows. These results suggest that integrating persistent execution with session-level provenance can improve the reliability of AI-assisted scientific workflows. Environment specification and full rerunnability remain weak for every evaluated system, ours included, so reproducibility is still an open problem for scientific agents. The system is available under the MIT license at \href{https://github.com/PKU-YuanGroup/OpenAI4S}{github.com/PKU-YuanGroup/OpenAI4S}.
AI agents are emerging as a practical way to run multi-step scientific workflows that interleave reasoning, tool use, and verification. Scaling such agentic science remains difficult because workflows are hard to observe and reproduce, many scientific tools and laboratory systems are not agent-ready, and execution trac...
Lin-Feng Zhang, Si-Heng Chen, Yu-Zhu Cai et al.· AI Plus· 0 citations
A multi-agent harness that represents each request as a persistent, artifact-grounded analysis run, and demonstrates that reliable bioinformatics automation depends not only on model capability, but also on the harness that governs planning, execution, memory, and evidence flow.
Qi Shao, Xin Zhang, Zhou-Yang Yuan et al.· 0 citations
OpenDiscoveryTrace, a public dataset of 558 complete AI scientific agent trajectories that captures how models reason, not just what they produce, is presented to support research on process-level evaluation, scientific agent auditing, and AI governance.
Summary Large language models (LLMs) increasingly write code, analyze data, and orchestrate scientific workflows. This can create reproducibility challenges because LLMs blur the boundary between how an analysis is built and what it depends on at run time. Guidelines for LLM-assisted science begin with a critical choic...
C. Dunn, Darrin T. Schultz, J. Musser· Patterns· 1 citation
Context: Large language model (LLM) agents are increasingly used as software and data-engineering assistants, yet evidence about locally deployable open-weight agents remains limited. Existing evaluations often emphasize textual responses or isolated code generation rather than the validity of complete engineering arti...
Jorge García-Carrasco, Javier Sanchis, A. Reina-Reina et al.· 0 citations
This manuscript reports the system design, verified capability boundary, deterministic software artifacts, and a reproducible evaluation protocol for BloClaw, an AI4S workstation built around a simple principle: a scientific agent should know what it can do, show how it did it, and state what remains unvalidated.
Yao Qin, Jin-Hua Pang, Xiao-Ming Zhang· bioRxiv· 0 citations