Claude Code transcript

Create install and launch scripts for: FeatureWiz Target OS: linux App metadata: {"id": "featurewiz", "name": "FeatureWiz", "description": "An automated feature selection and engineering library powered by the Minimum Redundancy Maximum Relevance (MRMR) algorithm. Enables users to select optimal features and create engineered variables with minimal code, supporting deep learning-based transformations.", "website": "https://github.com/AutoViML/featurewiz", "icon_url": "", "categories": ["Feature Engineering", "Machine Learning", "Data Science"], "tags": ["feature-selection", "feature-engineering", "machine-learning", "mrmr-algorithm", "automl", "data-science", "xgboost", "python"], "os_support": ["linux", "windows", "macos"], "requires_payment": false, "foss": true, "gh_repo": "https://github.com/AutoViML/featurewiz", "self_hostable": true, "requires_hardware": false, "package_managers": {"pip": "featurewiz", "conda": "featurewiz"}, "download_url": "https://pypi.org/project/featurewiz/"}

Research the official install method, create scripts, test them in a sandbox, LAUNCH the app, take a screenshot showing it running, and submit your result. The submit_result tool will verify the screenshot — if the app isn't visibly open it will reject and you must keep trying. Remember to delete the sandbox when done.