AgentBuild: A Framework for Constructing Scientific Agents Through Scientist-Authored Contracts
Researchers have introduced AgentBuild, a framework for constructing LLM-based scientific agents from structured contracts authored by domain scientists rather than through conventional AI training methods. The system was demonstrated on Rietveld refinement of X-ray diffraction data using the GSAS-II software package, where an agent was built and tested on a lithium lanthanum zirconium oxide (LLZO) dataset of varying scan quality. The approach aims to preserve scientific judgment in automated workflows by making the scientist's rubric—not the model—the durable, version-controlled asset.
A preprint submitted to arXiv proposes AgentBuild, a methodology for building scientific AI agents from a formal contract that the scientist authors, consisting of a versioned rubric, a difficulty-graded curriculum, and a curated external knowledge base. Unlike fine-tuning or reinforcement learning approaches, AgentBuild uses a rubric-driven judge to gate a meta-optimizer coding agent that modifies the scientific agent within declared boundaries, keeping the scientist's judgment central to the process. The framework was instantiated for Rietveld refinement—a standard crystallographic technique for analyzing X-ray diffraction data—using the GSAS-II software exposed through MCP and A2A interfaces. A blank-harness construction run was tested on a signal-to-noise ladder derived from LLZO scans, with a 4-hour scan identified as a frontier case where the agent's workflow scope was exceeded. Crucially, the rubric scores both fit quality and trajectory scope, so frontier failures are classified as contract failures rather than pattern-fitting failures. The authors argue that as base language models improve, re-running AgentBuild constitutes a re-tune rather than a full rebuild, with the scientist's contract remaining the stable, reusable artifact across model generations.
What's missing
The paper does not report quantitative benchmarks comparing AgentBuild-constructed agents against human experts or alternative agent-building methods (e.g., fine-tuning baselines) on the same Rietveld refinement tasks. No peer review or external validation is noted at this preprint stage.
What different sources said
- arXiv cs.AICenter
Fantastic Scientific Agents and How to Build Them: AgentBuild for Rietveld Refinement
Related
Gut Bacteria Enzyme Found to Break Down Heat-Processed Food Compounds, Producing Novel Biogenic Amines
Researchers have discovered that an enzyme in common gut bacteria can degrade N-epsilon-carboxymethyllysine (CML), a compound formed during thermal food processing, producing previously unknown biogenic amines. The enzyme, ornithine decarboxylase SpeC from enterobacteria, acts on CML and related modified lysine derivatives through a low-level 'underground' catalytic activity. This finding suggests a previously unrecognized communication axis between thermally processed dietary compounds and gut microbial physiology, with potential implications for host health.
Full-Length Gene Sequencing Reveals Two Distinct Bacterial Communities in Black-Legged Ticks Expanding Into Canada
Researchers used Oxford Nanopore full-length 16S rRNA gene sequencing to characterize the microbiome of Ixodes scapularis black-legged ticks collected in Nova Scotia, Canada, distinguishing between tick-adapted bacteria and environmentally acquired bacteria. The study comes as I. scapularis — the primary vector of Lyme disease — is rapidly expanding northward into Canada due to climate change. The findings suggest that environmentally derived bacteria in tick microbiomes are not mere contamination, which has implications for how tick microbiome data is collected and interpreted across surveillance studies.
Study Identifies Metabolic Link Between Cell Envelope Stress and Biofilm Formation in Bacteria
Researchers have discovered that the metabolite acetyl-CoA directly inhibits enzymes that degrade the bacterial signaling molecule c-di-GMP, connecting cell envelope biosynthesis stress to biofilm formation in Pseudomonas aeruginosa. The study found that sub-inhibitory concentrations of antibiotics targeting early peptidoglycan biosynthesis — but not other antibiotic classes — elevate c-di-GMP levels by reducing phosphodiesterase activity, with acetyl-CoA competing for the enzyme active site. Because the relevant enzyme domain is broadly conserved across bacterial species, this checkpoint mechanism may be widespread and could have implications for understanding antibiotic-induced biofilm responses.