Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add topprismdata/cultivating-ml-agent --skill kaggle-cognitive-cost-optimizationgit clone --depth 1 https://github.com/topprismdata/cultivating-ml-agentWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/topprismdata/cultivating-ml-agent/kaggle-cognitive-cost-optimization)<a href="https://agentmods.dev/skills/topprismdata/cultivating-ml-agent/kaggle-cognitive-cost-optimization"><img src="https://agentmods.dev/badge/skills/topprismdata/cultivating-ml-agent/kaggle-cognitive-cost-optimization/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/topprismdata/cultivating-ml-agent/kaggle-cognitive-cost-optimization"><img src="https://agentmods.dev/badge/skills/topprismdata/cultivating-ml-agent/kaggle-cognitive-cost-optimization.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00188 | $0.01718 |
| Opus 5 | $0.00094 | $0.00859 |
| Sonnet 5 | $0.00038 | $0.00344 |
| Haiku 4.5 | $0.00019 | $0.00172 |
Grade A, and why
kaggle-cognitive-cost-optimization scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 12d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 141 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Kaggle Cognitive Cost Optimization: A Decision Framework
Problem
Kaggle has a hidden cost: quota + time + complexity. Most agents:
- Submit 5+ redundant times → 0 score gain
- Build elaborate custom models → lose to 0.8 × simple public kernel
- Re-submit 10× for "safety" → no benefit
- Spend 3 weeks on a problem that 1 public fork solves in 2 hours
Core Insight: The 0.6×BEST_PUBLIC Rule
After 20+ competitions (M8 Sports, S6E1-7, PTCG, ROGII, NeuroGolf, etc.):
For any Kaggle competition: choosing the best public kernel + ≤2 days tuning achieves ≥80% of the optimal result. The remaining 20% requires 10× more effort for marginal gain.
Evidence:
| Competition | Public #1 Score | My Best Custom | Ratio |
|---|---|---|---|
| M8 Sports | 0.1390 | 0.1407 | 0.988 |
| S6E4 Irrigation | 0.6596 | 0.6691 | 0.986 |
| S6E5 F1 Pit Stop | 0.1237 | 0.1239 | 0.998 |
| SST | 0.8078 | 0.8124 | 0.994 |
| House Prices | 0.11750 | 0.1194 | 0.984 |
| Mean | — | — | 0.99 |
In 5+ competitions, custom work BEAT public by < 2% on average. Top 5% requires 10× more effort for 5% gain.
Context / Trigger Conditions
Use this skill when:
- First 48h of a new competition: spend ≥50% of time scanning top 10 public kernels
- Quota running low: prioritize 1 submission to best public > 3 to own variants
- Custom work plateauing: if your best OOF < public #1, the public is more likely the right answer
- Deadline approaching: stop tuning, submit the best combined
- Re-submit temptation: never re-submit within 48h of convergence start (see trueskill-simulation-competition-strategy)
Solution: The 3-Quota-First Strategy
For most competitions, allocate quota as:
Quota 1: Submit best public kernel EXACTLY (don't modify)
→ Establishes baseline. Validates 0.6×rule.
Quota 2 (optional): Submit public + minimal tuning (one new feature, one new param)
→ Tests if simple adaptation works.
Quota 3 (optional): Custom work if you have a fundamentally different angle
→ Only if evidence suggests the public is missing key insight.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 12d ago First seen · 141 lines · 188 tokens per session scan A f5c024494472
kaggle-cognitive-cost-optimization is a skill published in the GitHub repository topprismdata/cultivating-ml-agent (5 stars, last pushed 14d ago), licensed MIT. It adds 188 tokens to every session and 1,718 once invoked, about $0.0009 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
huggingface-hub
Hugging Face Hub CLI (hf) — search, download, and upload models and datasets, manage repos, query datasets with SQL, deploy inference endpoints, manage Spaces and buckets.
tensorboard
Visualize training metrics, debug models with histograms, compare experiments, visualize model graphs, and profile performance with TensorBoard - Google's ML visualization toolkit.
mlflow
Track ML experiments, manage model registry with versioning, deploy models to production, and reproduce experiments with MLflow - framework-agnostic ML lifecycle platform.
datachain-knowledge
Use whenever datasets, cloud storage buckets, or data pipelines are mentioned — creating, saving, querying, listing, exploring, deleting, or processing data in S3, GCS, Azure Blob, or local storage. Also use when running any script that may create datasets as a side effect. Maintains a knowledge base at dc-knowledge/…
prompt-scanner
A scanner for text sent to an AI agent, looking for prompt injection and jailbreak attempts. Prompt injection is text that tries to override an agent's instructions; a jailbreak tries to bypass its safety limits.
install-openviking-memory
Install and configure the OpenViking long-term memory plugin for OpenClaw via natural conversation. Once installed, the plugin automatically captures facts from chats and recalls relevant context before each reply (auto-capture + auto-recall, cross-session). Covers prerequisites, install through OpenClaw's plugin…