README.md

---
language:
  - en
license: mit
tags:
  - security
  - api-security
  - vulnerability-detection
  - code
  - owasp
task_categories:
  - text-classification
  - text-generation
size_categories:
  - 1K<n<10K
---

# API Vulnerability Dataset (10K)

A dataset of **10,000 API-specific vulnerability samples** used to fine-tune [harsharajkumar273/api-security-qlora](https://huggingface.co/harsharajkumar273/api-security-qlora) — a QLoRA adapter on CodeLlama-7b for automated API security analysis.

## Dataset Summary

Each sample contains a vulnerable or clean API endpoint code snippet paired with a structured security analysis covering vulnerability type, severity, CWE ID, and a remediated version.

## Language & Framework Distribution

| Language | Share | Frameworks |
|---|---|---|
| Python | 46% | Flask, FastAPI, Django |
| JavaScript | 25% | Express.js, NestJS |
| Java | 15% | Spring Boot |
| PHP / Go / Ruby / C# | 14% | Laravel, Gin, Rails, ASP.NET |

## Vulnerability Distribution

| Vulnerability | Samples | CWE |
|---|---|---|
| SQL Injection | 2,425 | CWE-89 |
| Mass Assignment | 1,307 | CWE-915 |
| Path Traversal | 943 | CWE-22 |
| IDOR | 860 | CWE-639 |
| Broken Authorization | 792 | CWE-285 |
| Command Injection | 600 | CWE-78 |

## Severity Breakdown

- **Critical (43%)**: RCE, SQLi, unauthorized admin access
- **High (41%)**: Data leaks, IDOR, authorization bypass
- **Medium / Clean (16%)**: XSS, input validation warnings, baseline clean samples

## Usage

This dataset was used to fine-tune the [api-security-qlora](https://huggingface.co/harsharajkumar273/api-security-qlora) adapter and is also consumed directly by the [API Security Scanner](https://github.com/harsharajkumar/api-security) rules engine.

## Credits

**CS6380 — API Security Project**
Authors: Siddhanth Nilesh Jagtap · Tanuj Kenchannavar · Harsha Raj Kumar