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 store-sales-darts-chronos-blendgit 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/store-sales-darts-chronos-blend)<a href="https://agentmods.dev/skills/topprismdata/cultivating-ml-agent/store-sales-darts-chronos-blend"><img src="https://agentmods.dev/badge/skills/topprismdata/cultivating-ml-agent/store-sales-darts-chronos-blend/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/store-sales-darts-chronos-blend"><img src="https://agentmods.dev/badge/skills/topprismdata/cultivating-ml-agent/store-sales-darts-chronos-blend.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 | $0.00247 | $0.01717 |
| Opus 5 | $0.00123 | $0.00859 |
| Sonnet 5 | $0.00049 | $0.00343 |
| Haiku 4.5 | $0.00025 | $0.00172 |
Grade A, and why
store-sales-darts-chronos-blend 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 9d 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 — 124 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Store Sales — darts + Chronos-2 Cross-Family Blend
The Core Insight (why this beats single models)
A single strong model hits a ceiling. Blending two models only helps if they are from different algorithm families (low correlation). Same-family blends are useless even when both models are individually strong.
| Blend | Correlation | Result |
|---|---|---|
| Chronos-2 v1 + Chronos-2 v2 (same family) | 0.997 | no gain (0.3939 → 0.3941) |
| Chronos-2 + darts-LightGBM (different families) | 0.997* | -0.009 gain (0.3939 → 0.3844) |
* Even at 0.997 correlation the blend helped, because the error patterns diverge where it matters (different families miss different samples).
The Two Routes
Route 1: AutoGluon Chronos-2 ensemble (neural foundation model)
TimeSeriesPredictorwith Chronos-2 (zero-shot + LoRA fine-tune) + Chronos-Bolt + DirectTabularnum_val_windows=5, local model paths to bypass HF download errors- Standalone LB: 0.39387
- See
autogluon-timeseries-strategyskill for Chronos-2 details
Route 2: darts LightGBM per-family (tree model, top-1 public method)
dartslibrary'sLightGBMModelwithoutput_chunk_length=1+predict(n=16)- darts handles recursive prediction AUTOMATICALLY — this is the critical advantage. Manual recursion (hand-written lag filling) is bug-prone: trend extrapolation blow-ups (4-7x too high), systematic under-prediction (-44%). darts' built-in recursion avoids all of these.
- Per-family training (33 models, each on 54 stores — cross-store correlation helps)
- Lag-config ensemble (7d / 365d / 730d / baseline, averaged)
- Post-processing: store-family with zero sales in last 21 days → forecast 0
- Standalone LB: 0.39953
from darts import TimeSeries
from darts.models import LightGBMModel
ts = TimeSeries.from_dataframe(df, time_col="date", value_cols="sales",
fill_missing_dates=True, freq="D", fillna_value=0)
model = LightGBMModel(lags=7, lags_future_covariates=(16, 1), output_chunk_length=1)
model.fit(series=[ts_store1, ts_store2, ...], # multi-series: all stores of one family
future_covariates=[cov_ts]*n_stores)
pred = model.predict(n=16, series=ts_store, future_covariates=cov_ts) # auto-recursive
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.
- 9d ago First seen · 124 lines · 247 tokens per session scan A 5a1adfa0452e
store-sales-darts-chronos-blend is a skill published in the GitHub repository topprismdata/cultivating-ml-agent (5 stars, last pushed 15d ago), licensed MIT. It adds 247 tokens to every session and 1,717 once invoked, about $0.0012 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 2026-09-03.
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…