README.md

---
license: cc-by-4.0
task_categories:
- text-classification
- token-classification
language:
- en
pretty_name: SmartUecDB
size_categories:
- 1K<n<10K
---

  
# SmartUecDB
## Description
SmartUecDB is a specialized, multimodal dataset designed for the high-fidelity detection of vulnerabilities in smart contract bytecode, with a primary focus on identifying **unchecked external calls**. The dataset curates a robust corpus of EVM-compatible execution traces and structural graph representations, enabling the training of Graph Neural Networks (GNN) to pinpoint insecure dispatch patterns and reentrancy vectors.

## Usage
You can start by downloading the primary dataset folder:

```python
import os
import requests
import tarfile
import shutil

FILE = "data.tar"
url = f"https://huggingface.co/datasets/Adson59/SmartUecDB/resolve/main/{FILE}"

def download_dataset(url, filename):
    response = requests.get(url)
    if response.status_code == 200:
        with open(filename, 'wb') as f:
            f.write(response.content)
        print(f"Successfully downloaded {filename}")
    else:
        print(f"Failed to download. Status code: {response.status_code}")

download_dataset(url, FILE)
```

```python
with tarfile.open("data.tar", 'r') as tar:
    tar.extractall(".")
```

## Structure
The following is the standard directory structure for the dataset, organized to support modular GNN training and inference pipelines:
```bash
SmartUecDB/
├── processed/          # cleaned and prepared rows
├── raw/                # original rows collected from the smartmaldb
├── split/              # Train/validation/test splits defined for reproducible model benchmarking
```

## Citation
If you use this dataset, cite:
```bibtex
@misc{smartuecdb25,
  author = {B. William},
  title = {SmartUecDB},
  year = {2026},
  publisher = {Hugging Face},
  url= {https://huggingface.co/datasets/Adson59/SmartUecDB},
}
```

---
license: cc-by-4.0
---