active · HolobiomicsLab
Magnolia — an AI agent that keeps what computational work teaches
A persistent-memory AI agent framework for computational research, with scientific tooling, a memory that is reviewed before it becomes a rule, and provenance by construction.
Also presented by the eDIAM team, Institut de Chimie de Nice
Magnolia helps a researcher retain and reuse the operational knowledge produced during computational work: which commands ran, which choices were made, what failed and why. It is used daily on the team's computational chemistry projects.
It wraps scientific tools (P2Rank, HADDOCK3, gnina, GROMACS, ORCA, Gaussian, xTB) on a workstation or a Slurm cluster, records runs, commands and outcomes as they happen, and promotes what it learns into version-controlled rules only with explicit agreement. Released publicly in September 2026 under the MIT licence with a Non-Military Clause; research software under active development.
- Showcase: Explore Magnolia
- Code: github.com/HolobiomicsLab/Magnolia-AI
- Preprint: arXiv:2610.27230