Dataset · Zaevlad
23.6K smart contract audit findings
Zaevlad/audit-findings-dataset
Raw smart contract vulnerability audit reports with descriptions, PoC code, and severity assessments.
What's inside
Raw semi-structured data rather than a ready-to-train dataset: requires deduplication, label normalization, and filtering before use. Contains 23,625 findings from smart contract security audits: title, description, proof-of-concept code, recommendation, and severity score. English language.
What you can do with it
- Classifying vulnerability severity and types after cleaning
- Generating remediation recommendations for audit findings
- Testing smart contract vulnerability detection systems
- Tasks
- Text classification, Text generation
- Languages
- English
- Size
- 10K–100K rows
- Downloads this month
- 445
- Likes
- 109
How to load it
from datasets import load_dataset
ds = load_dataset("Zaevlad/audit-findings-dataset")