MolAgent

Make predictions using a trained AutoMol model. Auto-discovers models from the registry. Use when the user wants to run inference on new molecules.

Plan and execute an AutoMol training pipeline. Analyzes dataset, auto-configures a plan, asks for approval, then executes — all in one invocation. Re-invoke to resume if interrupted. Use when the user wants to train a complete model from raw data.

Generate an interactive HTML dashboard from AutoMol evaluation results. Auto-discovers completed runs with evaluation data and renders Plotly.js charts. Use when the user wants to visualize training results.