What problem does it solve? Conducting scientific research involves many disconnected stages—forming hypotheses, designing methodologies, running computational analyses, and writing publication-ready papers. This Skill automates that entire pipeline using a multiagent AI system, so researchers can go from a dataset description to a formatted LaTeX manuscript without manually coordinating each step. ## Core Features & Use Cases - End-to-End Research Pipeline: Progress through four structured stages—data description, idea generation, methodology development, and results execution—with each stage producing version-controllable markdown outputs. - Publication-Ready Papers: Generate LaTeX papers formatted for specific journals (e.g., Journal.APS) with integrated figures and complete source files. - Flexible Hybrid Workflows: Mix manual and automated stages—provide your own hypothesis, methodology, or pre-computed results while automating the rest. - Use Case: A climate scientist describes a temperature anomaly dataset and available tools (pandas, scipy, sklearn), then runs the full pipeline to generate a research hypothesis, execute trend analysis with visualizations, and produce an APS-formatted paper. ## Quick Start Install denario with uv, set your LLM API key, then ask the AI to describe your dataset and run the full research pipeline to generate a paper.
