​

AI agent skills by research field

Research Agent Skills from Kalaris Labs combines scientific data tools with skills for literature review, academic writing, citations and publication. Researchers in academia can start with a focused bundle: ml-research, ai-research, biology-research, chemistry-research, medicine-research or physics-research. For example, run npx research-agent-skills install --bundle biology-research --project. Install an entire field category with npx research-agent-skills install --category <name>. The default research-essentials bundle covers writing, journal formats, literature review, ideation and figures across disciplines.

​

Life sciences, genomics and bioinformatics

Install life-sciences for workflows involving sequencing, single-cell analysis, biological networks and biomedical data. For example, scanpy explores single-cell RNA-seq data, scvi-tools covers probabilistic single-cell models, and pydeseq2 supports differential expression analysis. Add scientific-databases when the task needs external records and identifiers.

​

Chemistry, drug discovery and materials

Install chemistry-and-drug-discovery for molecule handling and computational chemistry. Start with rdkit for cheminformatics, deepchem for molecular machine learning and molecular-dynamics for simulation workflows. Materials researchers can add pymatgen.

​

Clinical and health research

Install clinical-and-health for research workflows involving medical images, biomedical datasets and clinical reporting. pydicom helps inspect DICOM data, pathml covers computational pathology, and clinical-reports helps structure research reports. Check institutional rules before sharing patient data with any agent or external service.

​

Physics, astronomy and earth science

Install physical-sciences for scientific computing in these fields. astropy handles astronomy data and coordinates, qiskit supports quantum circuits, pymatgen covers materials analysis, and geopandas supports geospatial research.

​

Psychology, social science and statistics

Start with research-essentials and data-science-and-ml. Use experimental-design for study plans, statistical-power for sample-size reasoning, statistical-analysis for analyses, and apa7 for APA-formatted manuscripts. For evidence synthesis, follow the systematic review guide.

​

AI and machine learning research

Install the relevant ml-training, ml-evaluation-and-safety, ml-inference-and-ops, or multimodal-and-emerging category. ml-paper-writing covers ML conference papers; evaluating-llms-harness addresses model evaluation; pytorch-fsdp2 covers distributed training. Use reproducibility-statement to document experimental details.

​

Lab automation and research operations

Install lab-automation for supported instruments and protocol systems. Start with opentrons-integration, pylabrobot, or protocolsio-integration, depending on your equipment and workflow. Review generated protocols before executing them in a lab. Browse every skill in the catalog. Missing your field? Request a skill. For paper discovery across fields, firecrawl-research-index queries Firecrawl’s hosted paper index. For questions over PDFs already in your lab, use paper-corpus-rag.