Skip to content
Back to search

nvidia-kaggle

NVIDIA Kaggle Plugin gives agents end-to-end Kaggle competition workflows through a single skill, nvidia-kaggle-skill. It can gather competition context, study public writeups and notebooks, reproduce kernels locally, submit to competitions, and manage Ka

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today
Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/nvidia--nvidia-kaggle/badge.svg)](/listings/nvidia--nvidia-kaggle)

HTML

<a href="/listings/nvidia--nvidia-kaggle"><img src="/api/listings/nvidia--nvidia-kaggle/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

**What this does:** Downloads a public Kaggle notebook along with the competition data and any dataset inputs it depends on, and lays them out in a local workspace that's ready to run. Handy when you want to start from an existing baseline instead of from scratch.

/nvidia-kaggle:nvidia-kaggle-skill Download and set up this notebook locally so I can run it: https://www.kaggle.com/code/cdeotte/titanic-wcg-xgboost-0-84688

<details> <summary>Show answer</summary>

Kernel Setup: cdeotte/titanic-wcg-xgboost-0-84688

Competition: Titanic
Language: R
GPU: not required

Kernel pulled:
- working/titanic-wcg-xgboost-0-84688.irnb

Competition data downloaded:
- input/titanic/train.csv
- input/titanic/test.csv
- input/titanic/gender_submission.csv

Note: cells 25-28 reference a private or deleted dataset and will not run. All other cells are reproducible.

Workspace:
- titanic-wcg-xgboost-0-84688/README.md
- titanic-wcg-xgboost-0-84688/create_symlinks.sh
- titanic-wcg-xgboost-0-84688/input/
- titanic-wcg-xgboost-0-84688/working/

To run:
cd titanic-wcg-xgboost-0-84688/working
jupyter notebook titanic-wcg-xgboost-0-84688.irnb

</details>

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

.claude-plugin/marketplace.json · Manifest · 23 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

  1. {
  2. "name": "nvidia-kaggle",
  3. "owner": {
  4. "name": "nvidia-kaggle maintainers"
  5. },
  6. "metadata": {
  7. "description": "Claude Code marketplace for the NVIDIA Kaggle plugin.",
  8. "version": "0.1.0"
  9. },
  10. "plugins": [
  11. {
  12. "name": "nvidia-kaggle",
  13. "source": {
  14. "source": "url",
  15. "url": "https://github.com/NVIDIA/nvidia-kaggle.git",
  16. "ref": "main"
  17. },
  18. "description": "Claude Code plugin for Kaggle competition workflows.",
  19. "category": "data-science",
  20. "tags": ["kaggle", "competition", "plugin"]
  21. }
  22. ]
  23. }
README.md · README · 266 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

  1. # NVIDIA Kaggle Plugin
  2. [![NVIDIA](https://img.shields.io/badge/NVIDIA-Verified-76B900?style=flat&logo=nvidia&logoColor=white)](https://github.com/nvidia/skillspector)
  3. [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](LICENSE)
  4. [![Kaggle API](https://img.shields.io/badge/Kaggle-API-20BEFF?style=flat&logo=kaggle&logoColor=white)](https://www.kaggle.com/docs/api)
  5. [![Agent Skills Spec](https://img.shields.io/badge/Agent%20Skills-Specification-blue?style=flat)](https://agentskills.io)
  6. [![Kaggle Grandmasters Playbook](https://img.shields.io/badge/Kaggle%20Grandmasters-Playbook-76B900?style=flat&logo=nvidia&logoColor=white)](https://developer.nvidia.com/blog/the-kaggle-grandmasters-playbook-7-battle-tested-modeling-techniques-for-tabular-data/)
  7. NVIDIA Kaggle Plugin gives agents end-to-end Kaggle competition workflows through a single skill, `nvidia-kaggle-skill`. It can gather competition context, study public writeups and notebooks, reproduce kernels locally, submit to competitions, and manage Kaggle datasets.
  8. The skill accepts whatever's natural — a competition slug, a Kaggle URL, a kernel ref, a writeup URL, or a local path — and picks the right script to run based on the request.
  9. ![Demo: asking Claude Code to summarize the top AIMO Progress Prize 2 solutions](demo/demo.gif)
  10. Examples of supported workflows include:
  11. | Workflow | What it does |
  12. |---|---|
  13. | Competition context | Fetches the overview, rules, evaluation, timeline, and dataset description. |
  14. | Solution writeups | Finds leaderboard writeups and summarizes the strongest approaches. |
  15. | Discussions | Indexes Kaggle discussions, searches them, and reads selected threads. |
  16. | Kernels | Indexes, queries, reads, and researches public Kaggle kernels. |
  17. | Kernel reproduction | Downloads a Kaggle notebook and its inputs into a local workspace. |
  18. | Submission | Pushes a Kaggle kernel submission and polls for the final result. |
  19. | Dataset upload | Creates or updates a Kaggle dataset from a local folder. |
  20. ## Requirements
  21. - Python 3.10 or newer.
  22. - An agent runtime with plugin or Agent Skills support, such as Codex, Claude Code, or another compatible harness.
  23. - A Kaggle account and `KAGGLE_API_TOKEN` for API-backed workflows, including dataset access, kernels, submissions, and dataset uploads.
  24. ## Examples
  25. ### Top Solution Writeups
  26. **What this does:** Scrapes the competition leaderboard, opens the top-N solution discussions, and summarizes the techniques each top team used. Useful when starting a new competition and you want to know what has historically won.
  27. ```bash
  28. /nvidia-kaggle:nvidia-kaggle-skill Get the top 3 solution writeups from the AI Mathematical Olympiad - Progress Prize 2 competition and summarize the key strategies.
  29. ```
  30. <details>
  31. <summary>Show answer</summary>
  32. ```text
  33. AIMO Progress Prize 2 - Top 3 Writeup Summary
  34. Rank Team Approach
  35. 1st NemoSkills (NVIDIA) Large-scale SFT + TIR fine-tune + model merge + TensorRT-LLM/FP8/ReDrafter
  36. 2nd imagination-research SFT + DPO on DeepSeek-R1-14B + lmdeploy W4KV8 + dual CoT/Code prompting
  37. 3rd Aliev Off-the-shelf DeepSeek-R1-14B AWQ + branching, duplication, and early-stop inference
  38. Key strategies:
  39. 1. Tool-integrated reasoning improved the winning team's majority-vote accuracy.
  40. 2. DPO shortened outputs, enabling more samples within the time budget.
  41. 3. Quantization, fast inference engines, and speculative decoding were critical.
  42. 4. Multi-level early stopping reduced wasted generation.
  43. 5. Dynamic time budgeting helped allocate samples across questions.
  44. 6. Strong inference design was enough for 3rd place without fine-tuning.
  45. Files written:
  46. - 1st_place_writeup.md
  47. - 2nd_place_writeup.md
  48. - 3rd_place_writeup.md
  49. - summary.md
  50. ```
  51. </details>
  52. ### Competition Overview And Dataset
  53. **What this does:** Fetches the competition's overview, rules, evaluation metric, timeline, and dataset description in one pass. This is usually the first thing to run for a new competition — it gives the agent the framing it needs before writing any code.
  54. ```bash
  55. /nvidia-kaggle:nvidia-kaggle-skill Fetch the competition overview and dataset description for the ARC Prize 2025 competition.
  56. ```
  57. <details>
  58. <summary>Show answer</summary>
  59. ```text
  60. ARC Prize 2025 - Competition Overview
  61. Type: Research competition - $1,000,000 total prize pool
  62. Goal: Build AI systems that efficiently learn new skills and solve novel abstract reasoning tasks.
  63. Metric: Percent correct predictions, with exact grid match required and 2 attempts per test output.
  64. Submission: JSON predictions for 240 unseen test tasks. Notebooks only, no internet, up to 12h runtime.
  65. Dataset files:
  66. - arc-agi_training_challenges.json
  67. - arc-agi_training_solutions.json
  68. - arc-agi_evaluation_challenges.json
  69. - arc-agi_evaluation_solutions.json
  70. - arc-agi_test_challenges.json
  71. - sample_submission.json
  72. Total size: 6.91 MB
  73. License: Apache 2.0
  74. Grid format: rectangular matrices of integers 0-9, up to 30x30.
  75. ```
  76. </details>
  77. ### Public Kernel Research
  78. **What this does:** Indexes the public kernels for a competition into a local cache, then queries and reads the highest-ranked ones to summarize the modeling approaches, feature engineering patterns, and common pitfalls. Useful for getting up to speed on what the community has already tried before you write your own baseline.
  79. ```bash
  80. /nvidia-kaggle:nvidia-kaggle-skill Research the top public kernels for the Home Credit Default Risk competition and summarize the modeling approaches used.
  81. ```
  82. <details>
  83. <summary>Show answer</summary>
  84. ```text
  85. Home Credit Default Risk - Top Public Kernels Research
  86. Top public kernels included:
  87. - Start Here: A Gentle Introduction - Will Koehrsen
  88. - Home Credit: Complete EDA + Feature Importance - Lathwal
  89. - LightGBM with Simple Features - Aguiar
  90. - Introduction to Manual Feature Engineering - Will Koehrsen
  91. - Feature Selection with Null Importances - olivier
  92. Modeling approaches:
  93. 1. LightGBM dominated competitive public kernels, usually with KFold CV, OOF averaging, and early stopping on AUC.
  94. 2. XGBoost pipelines used broad table aggregations and ratio features such as DAYS_EMPLOYED_PERC and EXT_SOURCE products.
  95. 3. Stacking combined tree models and gradient boosting models with a logistic regression meta-learner.
  96. 4. Deep learning underperformed GBDT approaches on this tabular dataset.
  97. 5. Feature engineering across bureau, previous applications, POS_CASH, credit card, and installments tables was central.
  98. 6. Bayesian optimization and null-importance filtering improved feature sets and model tuning.
  99. ```
  100. </details>
  101. ### Kernel Setup
  102. **What this does:** Downloads a public Kaggle notebook along with the competition data and any dataset inputs it depends on, and lays them out in a local workspace that's ready to run. Handy when you want to start from an existing baseline instead of from scratch.
  103. ```bash
  104. /nvidia-kaggle:nvidia-kaggle-skill Download and set up this notebook locally so I can run it: https://www.kaggle.com/code/cdeotte/titanic-wcg-xgboost-0-84688
  105. ```
  106. <details>
  107. <summary>Show answer</summary>
  108. ```text
  109. Kernel Setup: cdeotte/titanic-wcg-xgboost-0-84688
  110. Competition: Titanic
  111. Language: R
  112. GPU: not required
  113. Kernel pulled:
  114. - working/titanic-wcg-xgboost-0-84688.irnb
  115. Competition data downloaded:
  116. - input/titanic/train.csv
  117. - input/titanic/test.csv
  118. - input/titanic/gender_submission.csv
  119. Note: cells 25-28 reference a private or deleted dataset and will not run. All other cells are reproducible.
  120. Workspace:
  121. - titanic-wcg-xgboost-0-84688/README.md
  122. - titanic-wcg-xgboost-0-84688/create_symlinks.sh
  123. - titanic-wcg-xgboost-0-84688/input/
  124. - titanic-wcg-xgboost-0-84688/working/
  125. To run:
  126. cd titanic-wcg-xgboost-0-84688/working
  127. jupyter notebook titanic-wcg-xgboost-0-84688.irnb
  128. ```
  129. </details>
  130. ### Demo: Competition Strategy Brief
  131. This demo chains the skill's research workflows — overview, discussions, kernels, writeups — into a single strategy brief: the metric and constraints, the winning techniques (each linked to its source), a baseline→top score ladder, and a few plots. Give the skill a natural-language prompt:
  132. > Research the rogii-wellbore-geology-prediction Kaggle competition with the nvidia-kaggle skill and write me a strategy brief on what it takes to do well. Include the key public notebooks and discussions as links, and a few plots for insight. Save the brief to brief.md and put any plots under plots/.
  133. Sample output briefs:
  134. - [Codex run](https://github.com/daxiongshu/competition-brief-demo/blob/main/codex_rogii-wellbore-geology-prediction_002/brief.md)
  135. - [Claude run](https://github.com/daxiongshu/competition-brief-demo/blob/main/claude_rogii-wellbore-geology-prediction_003/brief.md)
  136. ## Quick Install
  137. Install the marketplace for your agent runtime.
  138. **Codex**
  139. ```bash
  140. codex plugin marketplace add https://github.com/NVIDIA/nvidia-kaggle.git
  141. ```
  142. Then open a Codex session, run `/plugins`, choose the `nvidia-kaggle` marketplace, and install `nvidia-kaggle`.
  143. **Claude Code**
  144. ```bash
  145. claude plugin marketplace add https://github.com/NVIDIA/nvidia-kaggle.git
  146. claude plugin install nvidia-kaggle@nvidia-kaggle --scope user
  147. ```
  148. ## Local Install
  149. **Codex**
  150. ```bash
  151. codex plugin marketplace add .
  152. ```
  153. Then open a Codex session, run `/plugins`, choose the local `nvidia-kaggle` marketplace, and install `nvidia-kaggle`.
  154. **Claude Code**
  155. ```bash
  156. claude plugin marketplace add .
  157. claude plugin install nvidia-kaggle@nvidia-kaggle --scope user
  158. ```
  159. **Other Harnesses**
  160. For any agent runtime that supports Agent Skills directly, copy the whole skill directory into that runtime's skills directory:
  161. ```bash
  162. cp -R skills/nvidia-kaggle-skill <your-skills-directory>/
  163. ```
  164. The copied directory must include `SKILL.md`, the workflow markdown files, and the `scripts/` directory.
  165. ## Credentials
  166. Kaggle API workflows require `KAGGLE_API_TOKEN`, expected to contain a KGAT token string. Keep it in your environment, `.env`.
  167. ## Evaluation
  168. See [BENCHMARK.md](BENCHMARK.md) for the skill evaluation summary, including evaluated agents, task coverage, metrics, and with-skill uplift against baseline runs.
  169. ## Development
  170. Run the default test suite:
  171. ```bash
  172. uv run pytest
  173. ```
  174. Run Claude Code marketplace install coverage:
  175. ```bash
  176. uv run pytest --run-claude-plugin tests/test_claude_plugin_install.py
  177. ```
  178. Run Kaggle integration tests only when credentials and network access are available:
  179. ```bash
  180. uv run pytest --run-integration
  181. ```
  182. Before a release-facing change:
  183. ```bash
  184. uv sync
  185. uv run pytest
  186. claude plugin validate .
  187. ```
  188. Contributor guidance is in [CONTRIBUTING.md](CONTRIBUTING.md).

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.