skilld
All skills
shepsci avatar

/kaggle

@256664c
by Shep Scientificshepsci/kaggle-skill95 stars
8

Unified Kaggle skill. Use when the user explicitly mentions Kaggle, kaggle.com, a Kaggle URL, Kaggle competitions, Kaggle datasets/models/notebooks, Kaggle forums/discussions/writeups, Kaggle benchmarks, hackathons hosted on Kaggle, Kaggle badges, or Kaggle account setup. Do not use for generic ML, GPU/TPU, notebook, dataset, benchmark, or data-science tasks unless the user clearly ties them to Kaggle.

Ask your Agent

Use this Skill: https://skilld.dev/gh/shepsci/kaggle-skill/kaggle

This session only. Nothing lands on disk.

modulescompetitionshackathonsreferencesepisode-endpoints.md

≈536 tokens on demand. Your agent reads this file only when SKILL.md points to it.

Episode Endpoints — Simulation & Agent Evaluation

Three MCP tools for retrieving simulation episode data — agent logs, replays, and per-submission episode listings. Useful when evaluating agent-based hackathon submissions or auditing simulation runs.

list_submission_episodes — ✅ PASS

List the episodes generated by a competition submission.

Parameters:

  • submissionId (int) — the Kaggle submission id

Use this as the entry point: pull the episode ids first, then fetch logs or replays per episode.

get_episode_agent_logs — 🔬 BAD_PROBE

Download per-agent log output for one episode.

Parameters:

  • episodeId (int)
  • agentIndex (int) — 0-based agent index within the episode

Status: Marked BAD_PROBE in the 2026-04-22 audit (test infrastructure issue rather than confirmed tool failure). The tool is documented and likely works; verify behavior before depending on it. Capture both success and error shapes when you call it.

from skills.kaggle.shared.mcp_client import mcp_call, resolve_token

token = resolve_token()
resp = mcp_call("get_episode_agent_logs", {
    "episodeId": 12345678,
    "agentIndex": 0,
}, token=token)

get_episode_replay — 🔬 BAD_PROBE

Download the replay payload for one episode.

Parameters:

  • episodeId (int)

Status: Same caveat as get_episode_agent_logs. Tool exists; full live behavior unverified. Treat the response shape as discoverable rather than documented until a clean probe lands.

Recommended pattern

# 1. Pull the submission's episodes
episodes = mcp_call("list_submission_episodes", {"submissionId": SID}, token=token)

# 2. For each episode, optionally pull replay + per-agent logs
for ep_id in extracted_episode_ids:
    replay = mcp_call("get_episode_replay", {"episodeId": ep_id}, token=token)
    for i in range(num_agents):
        logs = mcp_call("get_episode_agent_logs", {
            "episodeId": ep_id, "agentIndex": i,
        }, token=token)

When the BAD_PROBE endpoints return unexpected shapes, surface the raw response to the caller — a future audit may reclassify them.

Source: SKILL.md on GitHub

Third-party checks

1 alert16d5 checks · Risk SAFE
  • Gen Agent Trust Hub16d

    The 'kaggle' skill is a comprehensive and secure integration for Kaggle platform operations. It implements multiple security best practices, including credential masking, restrictive file permissions, and protection against token leakage to third-party sites. It also handles untrusted user-generated content from forums using explicit boundary markers to prevent prompt injection.

  • Socket16d

    No alerts

  • Snyk16d

    Risk: MEDIUM · 1 issue

  • Runlayer6mo

    28/45 files flagged

  • ZeroLeaks5mo

    Score: 93/100 · 2 sections analyzed

Provenance

Signed by skilld at 256664c. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub 2 months ago.

Activeupdated 3 months ago

Capability

homepage
https://github.com/shepsci/kaggle-skill
All 1 allowed tools
Bash Read WebFetch Grep Glob
Other metadata
compatibility
Python 3.11+, pip packages kagglehub>=1.0.0, kaggle>=2.2.3, kagglesdk>=0.1.33,<1.0, requests, python-dotenv. Optional: playwright for browser badges; kaggle-benchmarks for local benchmark task authoring. The competitions module's SPA-scraping steps assume Playwright MCP tools are provided by the host agent; the skill itself does not bundle them.
metadata
{
  "author": "shepsci",
  "version": "2.4.0",
  "primaryEnv": "KAGGLE_API_TOKEN",
  "openclaw": {
    "requires": {
      "bins": [
        "python3",
        "pip3"
      ],
      "env": [
        "KAGGLE_API_TOKEN"
      ]
    }
  }
}

README badge

README badge for shepsci/kaggle-skill