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 responsible-genai-hackweek/nicer-nisar --skill nisar-snow-traingit clone --depth 1 https://github.com/responsible-genai-hackweek/nicer-nisarWrote 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/responsible-genai-hackweek/nicer-nisar/nisar-snow-train)<a href="https://agentmods.dev/skills/responsible-genai-hackweek/nicer-nisar/nisar-snow-train"><img src="https://agentmods.dev/badge/skills/responsible-genai-hackweek/nicer-nisar/nisar-snow-train/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/responsible-genai-hackweek/nicer-nisar/nisar-snow-train"><img src="https://agentmods.dev/badge/skills/responsible-genai-hackweek/nicer-nisar/nisar-snow-train.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 | $0.00097 | $0.02042 |
| Opus 5.5 | $0.00039 | $0.00817 |
| Sonnet 5.5 | $0.00019 | $0.00408 |
| Haiku 4.5 | $0.00010 | $0.00204 |
Grade A, and why
nisar-snow-train 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 2mo 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 — 156 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Training and freezing a model
Workflow A in CLAUDE.md. Offline, iterative, run once, then frozen. Unlike the prediction path this one has real choices in it, and most of them are choices about not fooling yourself.
matched (InSAR pair, LiDAR flight pair)
→ stream GUNW → build_feature_table → attach_target → apply_row_filters
→ cross_validate + robustness_report → pick estimator → freeze → pin model_id
Entry point: python scripts/run_train.py. Target aggregation is separable and
slow: python scripts/build_targets.py dsd caches it first.
When to retrain
- A new matched pair becomes usable, or you extend to more of the six.
FEATURIZER_VERSIONorANCILLARY_VERSIONbumps.model.load()will refuse the old artifact — that refusal is the design working, not a bug to route around.FEATURE_COLUMNSchanges. It is frozen and ordered; changing it mints a newmodel_idby construction.
Never edit a frozen model's card to make it load. Refit and pin a new id.
Choosing pairs
C.MATCHED_PAIRS holds the six 12-day InSAR pairs that land within a week of a
LiDAR flight on both ends. C.DEFAULT_PAIR_INDEX = 5 is the best match
(20260207→20260219, track 077, offsets 0 / +3).
Temporal offset is unmodelled error in the target, not bookkeeping. A ±7 day gap
means the LiDAR interval and the InSAR interval do not coincide — snow accumulated
or settled in between, and that difference lands in dsd_mean as if it were signal
the radar should have seen. Either weight matches by abs(off_ref) + abs(off_sec)
or restrict to the tight ones, and say which you did in the card.
Three November pairs predate the first flight (20251203) and are unusable. Track 149 has a contiguous run: 20251226 → 20260107 → 20260119 → 20260131 → 20260212.
Holdout — the decision that determines whether the score means anything
Hold out whole pairs, not pixels, and not spatial blocks if you can avoid it. Random pixel splits leak outright; neighbouring 80 m cells are near-duplicates.
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.
- 2mo ago First seen · 156 lines · 97 tokens per session scan A bc06ae1c3387
nisar-snow-train is a skill published in the GitHub repository responsible-genai-hackweek/nicer-nisar (2 stars, last pushed 2mo ago), licensed BSD-3-Clause. It adds 97 tokens to every session and 2,042 once invoked, about $0.0004 per session on Opus 5.5. 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
analyze-project
A read-only guide for understanding a deep-learning code repository, including its model structure, training and prediction entry points, and configuration links.
torchdrug
Builds and troubleshoots TorchDrug 0.2.1 workflows for molecular graphs, property prediction, self-supervised pretraining, molecule generation, retrosynthesis, protein representation learning, and knowledge graph reasoning. Use when code imports torchdrug or needs its datasets, models, tasks, or Engine.
arboreto
Infers candidate gene regulatory networks from bulk or single-cell expression data using AertsLab Arboreto GRNBoost2 and GENIE3. Use for transcription factor-target association ranking, compatible Dask execution, sparse expression inputs, and network stability checks.
deepspot-m
Generates transcriptome-wide virtual spatial transcriptomics from H&E histology with DeepSpot-M. Used for predicted log1p-CPM expression from 224x224 tiles at about 20x, querying the released protein-coding gene panel by symbol, and whole-slide prediction after resolution-aware tiling with histolab.
pyhealth
Builds and validates PyHealth clinical machine-learning pipelines for EHR, signals, imaging, and medical codes. Use for PyHealth dataset loading, MIMIC-III/IV, eICU or OMOP prediction tasks, patient-level evaluation, mortality/readmission/length-of-stay modeling, medication recommendation, sleep staging, Trainer…
nemo-mbridge-perf-expert-parallel-overlap
Validate and use MoE expert-parallel communication overlap in Megatron-Bridge, including overlapmoeexpertparallelcomm, delaywgradcompute, and flex dispatcher backends such as DeepEP and HybridEP.