{"id":"11307e90-9bb6-401f-92ea-543dc32beabc","slug":"kaggle-dominator","kind":"skill","name":"kaggle-dominator","summary":"Evidence-driven Kaggle strategy for coding agents: trustworthy validation, BEST_KNOWN protection, resource budgets, Pareto selection, and safe submission gates.","owner":"KapaSique","repoUrl":"https://github.com/KapaSique/kaggle-dominator","license":"MIT","licenseTier":"permissive","trust":"community","stars":3,"match":null,"fewStars":true,"format":"skill","createdAt":"2026-06-16T08:24:39.000Z","addedAt":"2026-10-22T01:19:05.450Z","logoUrl":"https://avatars.githubusercontent.com/u/82393399?u=d4354b0b4339241eceae604221c5ec2d372715a6&v=4","linkOnly":false,"collection":null,"skillPath":null,"categories":["developer-tools","data-analytics"],"tags":["ai-agent","anthropic","claude","claude-code","data-science","kaggle","kaggle-competition","llm"],"runtime":"binary","local":true,"cloudCapable":false,"cloudCapability":"unknown","cloudReadyOnPlatform":false,"needsApiKey":false,"reviewed":false,"syncedAt":"2026-10-26T03:50:09.912Z","sample":false,"url":"https://conduits.dev/tools/kaggle-dominator","language":"Python","transports":["stdio"],"install":{},"requirements":{"env":[],"upstreamMayCharge":false},"tools":null,"readme":"# Kaggle Dominator\n\nAn evidence-driven strategy and execution skill for legitimate Kaggle competition\nwork across tabular, computer vision, NLP/audio, simulation and agent ladders,\ncode competitions, and judge-scored hackathons.\n\nIt is the competitive brain: live reconnaissance, validation design,\n`BEST_KNOWN` protection, experiment portfolios, ensembling, campaign focus, and\nsubmission decisions. Pair it with a separate `kaggle` infrastructure skill for\nauthentication, downloads, kernels, and API plumbing.\n\n## What v3 adds\n\n- Universal routing across all major Kaggle formats.\n- Portfolio states (`ACTIVE`, `MONITOR_ONLY`, `PAUSED`, `CLOSED`) and explicit\n  budgets for attention, compute, submissions, storage, and deadline slack.\n- A candidate queue with exploit/explore/audit lanes and stop/continue gates.\n- Fresh-score, metric-direction, submission-availability, and provenance checks.\n- Clear authority boundaries for submissions and public actions.\n- Compliance, attribution, reproducibility, and anti-leakage safeguards.\n- Learning checklists, public-resource guidance, an account scorecard, measured\n  cross-domain lessons, and bounded autonomous loops.\n- Contract tests and Codex discovery metadata.\n\n## Core loop\n\n```text\nlive recon\n  → front card and resource budget\n  → reproduce strongest compliant baseline\n  → trustworthy validation / arena\n  → candidate batch with cheap gates\n  → full measurement and diversity analysis\n  → submission decision with authority + quota checks\n  → fresh score, ledger update, next decision\n```\n\nA candidate never replaces `BEST_KNOWN` until the correct real metric confirms it.\nIf validation disagrees with the leaderboard or ladder, repair validation before\ncontinuing to tune.\n\n## References\n\n| Need | Reference |\n|---|---|\n| Multiple fronts and resource focus | [`campaign-control.md`](references/campaign-control.md) |\n| Cross-domain competitive process | [`grandmaster-playbook.md`](references/grandmaster-playbook.md) |\n| Measured account history | [`scorecard.md`](references/scorecard.md) |\n| Tabular | [`tabular.md`](references/tabular.md) |\n| CV, NLP, audio, signals | [`deep-learning.md`](references/deep-learning.md) |\n| Simulation and agents | [`simulation.md`](references/simulation.md) |\n| Code competitions and hackathons | [`code-and-hackathon.md`](references/code-and-hackathon.md) |\n| Autonomous bounded loops | [`autonomous.md`](references/autonomous.md) |\n| Self-curation | [`self-improvement.md`](references/self-improvement.md) |\n| ML craft | [`learning-craft.md`](references/learning-craft.md) |\n| Learning sources and attribution | [`resources.md`](references/resources.md) |\n| Winning solution patterns | [`winning-solutions.md`](references/winning-solutions.md) |\n| Tools and communities | [`arsenal.md`](references/arsenal.md) |\n\n## Install\n\nCodex project-local installation:\n\n```bash\ngit clone https://github.com/KapaSique/kaggle-dominator.git \\\n  .agents/skills/kaggle-dominator\n```\n\nCodex user installation:\n\n```bash\ngit clone https://github.com/KapaSique/kaggle-dominator.git \\\n  ~/.agents/skills/kaggle-dominator\n```\n\nInvoke it explicitly as `$kaggle-dominator`, or let it trigger on competitive\nKaggle strategy requests. It deliberately does not trigger for badge collection,\naccount configuration, or CLI installation alone.\n\n## Verify\n\n```bash\npython3 -m unittest -v tests/test_skill_contract.py\npython3 /path/to/skill-creator/scripts/quick_validate.py .\nbash -n scripts/*.sh\npython3 -m py_compile scripts/*.py tests/*.py\n```\n\n## License\n\n[MIT](LICENSE)\n"}