Denario
An AI multi-agent system that acts as a scientific research assistant — generating ideas, checking the literature, developing research plans, writing and executing code, making plots, and drafting and reviewing scientific papers.
Install in seconds: pip install denario-mcp
A modular research pipeline
Each module handles a specific task — or chain them together for an end-to-end scientific analysis.
Idea
Generates research ideas through iterative refinement — an “idea maker” and an “idea hater” critique and improve each concept.
Literature
Validates novelty by classifying ideas as new or not new and retrieving relevant papers through semantic search.
Planning & Control
Deep research orchestration via the Cmbagent backend, enabling end-to-end analysis without human intervention.
Writing
A multi-agent workflow that produces complete scientific papers with methods, results, and supporting materials.
Execution
Writes and runs code, makes plots, and turns a research plan into reproducible results.
GUI & MCP
Drive Denario from a friendly app, or straight from Claude through the hosted MCP server.
Run it from Claude, securely
Denario is available as a hosted MCP server. Install the lightweight client, create an API key, and the research tools appear inside Claude. The heavy pipeline runs on our infrastructure; your project files are created on your own machine.
- 1
Create an API key
Sign in and generate a key on the dashboard.
- 2
Install the client
pip install denario-mcp
- 3
Add to Claude
Paste the MCP config with your key + server URL.
- 4
Ask Claude
Run denario_setup, then denario_idea.
Proven across disciplines
A Denario-generated paper was accepted for publication at the Open Conference of AI Agents for Science 2025 — the first open conference with AI as primary authors.