← ClaudeAtlas

kg-startlisted

Bootstrap a Kaggle competition workspace end to end — recon, data download, workspace init, EDA, adversarial validation, frozen folds, metric implementation, baseline model and a first validated submission file.
TranBaDat2607/claude-code-kaggle-grandmaster · ★ 2 · AI & Automation · score 78
Install: claude install-skill TranBaDat2607/claude-code-kaggle-grandmaster
# /kg-start — bootstrap a competition Run kgkit with the Bash tool, as `PYTHONPATH="${CLAUDE_PLUGIN_ROOT}" python -m kgkit <command>`: this command pre-approves that form for the kgkit calls it makes. Submitting, pushing or publishing still asks (so does a PowerShell `$env:` assignment). Arguments: `$ARGUMENTS` (competition slug, optional workspace directory; default: current directory). kgkit: use the absolute path from the session context ("kgkit home"), else `$KGKIT_HOME`, else `${CLAUDE_PLUGIN_ROOT}`. Invoke as `PYTHONPATH=<kgkit home> python -m kgkit <cmd>` (PowerShell: `$env:PYTHONPATH="<kgkit home>"; python -m kgkit <cmd>`). Use `python3` if `python` is not the right interpreter on this machine. Follow the `grandmaster-playbook` and `competition-recon` skills. Steps: 1. **Preflight**: `kaggle --version` (offer `pip install kaggle` if missing). Check auth by running `kaggle competitions files <slug>`; on 401/403 tell the user to create an API token and/or accept the competition rules on the website — do not proceed with downloads until fixed. Never print credentials. 2. **Workspace**: create/enter the directory; `git init` if not a repo. 3. **Recon** (in parallel with the download): launch the `competition-analyst` agent for the slug (and optionally `solution-researcher` for prior art). Meanwhile: 4. **Data**: `kaggle competitions download <slug> -p data/` and extract. List files and sizes. 5. **Init**: `python -m kgkit init <slug> --metric <m> --task <