Scenarios
Turn research tasks into workflows your agent can execute.
Start from your current research scenario to find the Skills to install, ready-to-copy prompts, expected outputs, and next actions—from topic framing and literature search to pre-submission checks.
3 scenarios
From Question to Evidence
Turn ideas, literature, study design, and analysis into reviewable research judgment.
literature-evidence
Literature to evidence
Move from search, full-text reading, and citation checks to evidence tables that support research judgment.
You will get
study-design
Design and statistics
Clarify hypotheses, variables, randomization, power, statistical tests, and reporting before data collection.
You will get
data-analysis
Data analysis pipeline
Connect exploratory analysis, cleaning, modeling, statistical interpretation, and reproducible execution across tabular and scientific data.
You will get
4 scenarios
Computational Research and Domain Tools
Cover AI/ML experiments, life science, proteins and molecules, materials, geospatial, and physical science workflows.
ai-experiments
AI and model experiments
Discover scientific datasets, train models, optimize GPU workloads, iterate benchmarks, and record results for AI4Science, LLM, ML, and systems research.
You will get
life-science
Life science data
Cover single-cell, bulk RNA-seq, pathway enrichment, biomedical records, and public biological databases, connecting domain tooling to publishable results.
You will get
chem-materials
Proteins, molecules, materials
Organize structure prediction, molecular descriptors, docking, drug discovery, and materials tools into screening, modeling, interpretation, and reporting workflows.
You will get
physical-science
Geo and physical science
Handle remote sensing/GIS, astronomy FITS, quantum or physical simulation, and engineering optimization beyond writing-centric workflows.
You will get
3 scenarios
Lab, Compute, and Reusable Workflows
Turn lab platforms, remote compute, model endpoints, and repeated team workflows into executable, handoff-ready workbenches.
lab-workflow
Lab automation
Connect protocols.io, Benchling, Opentrons, Pylabrobot, and experiment logs into one traceable execution chain.
You will get
compute-infra
Compute and environments
Choose execution environments across local machines, SSH/Slurm, Modal, cloud GPUs, and model endpoints while tracking dependencies, resources, cost, and cleanup.
You will get
custom-workflow
Custom research agents
Observe repeated workflows, compose existing skills, write new skills, record project memory, and build reusable agents for the team.
You will get
2 scenarios
Publication and Communication
Move results into manuscripts, reviewer responses, data availability statements, figure panels, slides, and posters.
publication
Manuscript and submission
Cover the publication half of the journey: argument structure, English polish, citation and data availability, pre-submission review, and point-by-point responses.
You will get
research-visuals
Figures and presentations
Unify claims, evidence, layout, venue formats, and talk narrative into deliverable research visuals.
You will get