A workflow for training and using several machine-learning models on fMRI scans. It covers preparing data, pretraining, fine-tuning, making predictions, and comparing model results.

Use it to predict age, gender, health traits, or disease; classify brain states; retrieve similar scans; and benchmark models on datasets such as HCP1200, ABCD, and UK Biobank.

It brings multiple fMRI model types and research datasets into one workflow, reducing the need to manage separate training setups.

neurostorm — Skill for Claude Code and Codex by CUHK-AIM-Group

Install

Getting it into your agent

It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.

Clone the repo

git clone --depth 1 https://github.com/CUHK-AIM-Group/NeuroClaw

agentmods

npx agentmods add skills/cuhk-aim-group/neuroclaw/neurostorm

Made for: Claude Code, Codex.

Wrote 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.

README.md

[![agentmods](https://agentmods.dev/badge/skills/cuhk-aim-group/neuroclaw/neurostorm/github.svg)](https://agentmods.dev/skills/cuhk-aim-group/neuroclaw/neurostorm)

Your own site

<a href="https://agentmods.dev/skills/cuhk-aim-group/neuroclaw/neurostorm"><img src="https://agentmods.dev/badge/skills/cuhk-aim-group/neuroclaw/neurostorm/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.

Your own site · 80×15

<a href="https://agentmods.dev/skills/cuhk-aim-group/neuroclaw/neurostorm"><img src="https://agentmods.dev/badge/skills/cuhk-aim-group/neuroclaw/neurostorm.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 255 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 3,629 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits

NVIDIA SkillSpector warn 7 Sept 2026

SkillSpector: 3 findings, up to medium

These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →

medium Data Exfiltration · line 70
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Fix: Verify the destination URL is trusted and necessary. Remove or replace with documented APIs. Ensure no secrets, tokens, or PII are transmitted.
medium MCP Rug Pull · line 86
Docker image references without a specific tag (:latest is implicit) or digest (@sha256:...) can be silently replaced by a malicious image.

Fix: Pin the image: image:tag or image@sha256:abc123
low Tool Misuse · line 86
Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Fix: Validate all tool parameters against an allowlist. Reject dangerous parameter values (shell=True, --force, -rf /) and use safe defaults.

How audits are shown

Origin original No closer match found in the catalogue.

Token cost

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	$0.00255	$0.03629
Opus 5	$0.00128	$0.01814
Sonnet 5	$0.00051	$0.00726
Haiku 4.5	$0.00026	$0.00363

Measured 9d ago against content hash dde6566b4173, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-12, from the pricing page.

Security

Grade A, and why

neurostorm 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.

skills/neurostorm/SKILL.md · 373 lines

How it starts

The opening of the file, as written. The whole thing — 373 lines — stays where its author put it; the contents beside it link to each section on GitHub.

NeuroSTORM Skill

Overview

neurostorm-skill wraps the unified NeuroSTORM fMRI platform (CUHK-AIM-Group), which, as of the 2026-05-08 release, ships 8 model implementations under a single training/fine-tuning entry point. Use this skill for the full lifecycle: data download, preprocessing, pretraining, fine-tuning, and inference.

Supported models (8)

Model	Input type	Graph?	Backbone
`neurostorm`	voxel (4D)	No	Mamba-SSM
`swift`	voxel (4D)	No	Swin 4D Transformer
`braingnn`	FC graph (2D)	Yes	GNN
`bnt`	FC matrix (2D)	No	Transformer
`lggnn`	ROI + FC	Yes	Learnable GNN
`combraintf`	FC matrix (2D)	No	Community-aware Transformer
`ibgnn`	FC graph (2D)	Yes	Interpretable GNN
`brainnetcnn`	FC matrix (2D)	No	CNN

Supported tasks

ID	Task
1	Age & Gender Prediction
2	Phenotype Prediction
3	Disease Diagnosis
4	fMRI Retrieval
5	Task fMRI State Classification

Supported datasets: HCP1200, ABCD, UKB, Cobre, ADHD200, HCPA, HCPD, UCLA, HCPEP, HCPTASK, GOD, NSD, BOLD5000.

Dual data formats: PT (faster random access, larger disk) and H5 (compact, scales to large cohorts). Choose at preprocessing and at training via --output_format / --data_format.

Installation

Use the upstream requirements.txt + set_env.sh flow (Python 3.11, CUDA 12.8, PyTorch 2.7.1).

# 1. Clone and enter
git clone https://github.com/CUHK-AIM-Group/NeuroSTORM.git
cd NeuroSTORM

# 2. Create and activate env
conda create -n neurostorm python=3.11
conda activate neurostorm

# 3. Auto-detect conda + CUDA paths, set TORCH_CUDA_ARCH_LIST
source ./set_env.sh

# 4. Core dependencies
pip install -r requirements.txt
pip install "setuptools<81"               # pytorch-lightning 1.9.4 compat
pip install "transformers<=4.39.3"        # mamba-ssm compat

# 5. Graph-based models (BrainGNN / LG-GNN / IBGNN)
pip install torch-geometric
pip install torch-scatter torch-sparse -f https://data.pyg.org/whl/torch-2.7.0+cu128.html

# 6. FC-based models (BNT / BrainNetCNN / Com-BrainTF)
pip install scikit-learn pandas h5py deepdish

# 7. Mamba-SSM (NeuroSTORM only)
bash scripts/install_mamba.sh
#   or manually: causal-conv1d v1.5.0.post8, mamba v2.2.2
#   both built with TORCH_CUDA_ARCH_LIST matching your GPU (12.0 Blackwell,
#   9.0 H100, 8.9 4090, 8.6 3090, 8.0 A100)

Read the full file on GitHub · 373 lines

Changes

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 · 373 lines · 255 tokens per session scan A dde6566b4173

Subscribe to this mod's changes

neurostorm is a skill published in the GitHub repository CUHK-AIM-Group/NeuroClaw (85 stars, last pushed 6d ago), licensed MIT. It adds 255 tokens to every session and 3,629 once invoked, about $0.0013 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.

Data and AI Scientific

arboreto

Infer gene regulatory networks (GRNs) from gene expression data using scalable algorithms (GRNBoost2, GENIE3). Use when analyzing transcriptomics data (bulk RNA-seq, single-cell RNA-seq) to identify transcription factor-target gene relationships and regulatory interactions. Supports distributed computation for…

K-Dense-AI/scientific-agent-skills · 66 tokens

pyhealth

Build clinical/healthcare deep-learning pipelines with PyHealth — loading EHR/signal/imaging datasets (MIMIC-III/IV, eICU, OMOP, SleepEDF, ChestXray14, EHRShot), defining tasks (mortality, readmission, length-of-stay, drug recommendation, sleep staging, ICD coding, EEG events), instantiating models (Transformer…

K-Dense-AI/scientific-agent-skills · 216 tokens

torchdrug

Build and troubleshoot 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.

K-Dense-AI/scientific-agent-skills · 61 tokens

deepspot-m

Generate transcriptome-wide virtual spatial transcriptomics from H&E histology with DeepSpot-M. Use when you need spatial gene expression in log1p-CPM for 224x224 tiles at about 20x, want to query protein-coding genes by symbol instead of a fixed panel, or want to run prediction across a whole slide after tiling with…

K-Dense-AI/scientific-agent-skills · 80 tokens

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.

NVIDIA/skills · 56 tokens

pick-a-pii-model

Select an on-device OpenMed PII model from the committed registry by language, runtime format, and size budget, then require recall validation before deployment. Use when an agent must choose a local PII detector for CPU, Apple Silicon, or a mobile export without relying on live model discovery.

maziyarpanahi/openmed · 64 tokens