Add Atomcraft - AI platform for materials discovery

[fridayowl]

Adding Atomcraft to materials-discovery category.\n\nAtomcraft is an AI operating system for accelerated materials discovery with crystal structure generation, 11 ML property prediction models, synthesis feasibility engine, DFT simulation orchestration (VASP/QE/CP2K/LAMMPS), active learning loop, and a full React frontend.

[mergify]

Tick the box to add this pull request to the merge queue (same as @mergifyio queue).

• Queue this pull request

[fridayowl]

@Mergifyio queue

[mergify]

queue

☑️ Command disallowed due to command restrictions in the Mergify configuration.

Details

• sender-permission >= write

[Irratzo]

Dear @fridayowl,

Thanks for the proposed addition. Sorry for the delay.

I am a bit skeptical about adding your project. It seems entirely vibe-coded to me. All project commits on one day for a fully functional end-to-end materials discovery platform that encompasses all features that other projects in this list focus on one by one. To the point that I questioned the advertised functionality.

So, I tested it out. I had to fix the DB initialization in order to get the backend to work. That led to my PR #1 to your project. After that, the backend demos worked. Then I checked the frontend GUI, with the now populated database. The frontend GUI runs locally. I tried out all functionalities in there. None worked. To conclude the test, from my side: The backend works (with the PR1), the frontend does not.

Here is my verdict and my proposal. At this point, the project does not seem mature enough to be included in this list, for what it advertises vs. what actually works. However, if you fix the following points, I will reconsider adding the project to the list.

• Your GitHub profile features an ORCiD ID (https://orcid.org/0009-0001-1328-0190), but it does not exist at time of writing: "There's no displayable data for this record". Get a real one or fix the broken link.
• The project's Zenodo DOI in its CITATION.cff file does not exist at time of writing (10.5281/zenodo.14859548). I don't know whether its hallucinated. Get a real one or fix the broken DOI or remove it.
• Merge my PR #1 into your project or otherwise fix and document the database initialization step so the backend quickstart and demos work.
• Clearly state in the project documentation, for instance with a roadmap, what features are functional yet (and how to get them working comprehensively), and what features are as yet in planning phase.

Thank you and all the best.