Skip to content
KitploitKITPLOIT
ToolsExploitsBlog
Log in
Einreichen
ToolsExploitsBlog
Einreichen

Hacking-, PenTest- und Cybersicherheits-Tools für Ihr Sicherheitsarsenal!

Kitploit ist ein Verzeichnis von Hacking-, Cybersicherheits- und Pentesting-Tools. Entdecken Sie die neuesten Projekt-Updates, um Schwachstellen zu finden, Systeme zu analysieren, Tests zu automatisieren und Ihre Sicherheit zu stärken.

FeedsKontaktDatenschutz© 2026 Kitploit

Tool-Verzeichnis

Kategorien

Alle Kategorien anzeigen
Loading categories
WAInjectBench — Benchmarking prompt injection detections for web agents. | Kitploit
Tools/GitHubGitHub/norrrrrrr-lyn/wainjectbench
Vulnerability AnalysisWeb SecurityMachine LearningPapers & ResearchLearning & EducationAI SecurityAdversarial Attack
GitHubnorrrrrrr-lyn/wainjectbench

WAInjectBench

Benchmarking prompt injection detections for web agents.

Repository anzeigen
23594vor 4 MonatenVon Kitploit geprüft
Teilen
Inhalt in der angeforderten Sprache nicht verfügbar. Englische Version wird angezeigt.

WAInjectBench

WAInjectBench is a comprehensive benchmark for prompt injection detection in web agents.
It covers 6 types of attacks, across two modalities: text and image.


📂 Dataset Structure

  • data/
    • text/
      • benign/ → 4 categories, stored as JSONL files
      • malicious/ → 8 attack types, stored as JSONL files
    • image/
      • benign/ → 2 categories, stored in subfolders
      • malicious/ → 7 attack types, stored in subfolders

⚙️ Installation

Clone the repo and create the environment:

git clone https://github.com/Norrrrrrr-lyn/WAInjectBench.git
cd WAInjectBench
conda env create -f environment.yml
conda activate wainjectbench

🚀 Evaluation

WAInjectBench supports two evaluation pipelines: text-based detection and image-based detection.

🔹 Text-based Detection

python main_text.py \
  --data_dir [path to text dataset] \
  --detector [detector name] \
  --result_dir [output path] \
  --gpu [gpu id]

Available detectors: ["kad", "promptarmor", "embedding-t", "promptguard", "datasentinel", "ensemble"]

PromptArmor → requires OPENAI_API_KEY as environment variable.

DataSentinel →

git clone https://github.com/liu00222/Open-Prompt-Injection.git

Download the pretrained model into: WAInjectBench/Open-Prompt-Injection/DataSentinel_Models Set the directory and model path in detector_text/datasentinel.py.

🔹 Image-based Detection

python main_image.py \
  --data_dir [path to image dataset] \
  --detector [detector name] \
  --result_dir [output path] \
  --gpu [gpu id]

Available detectors: ["gpt-4o-prompt", "llava-1.5-7b-prompt", "jailguard", "embedding-i", "llava-1.5-7b-ft", "ensemble"]

GPT-4o-Prompt → requires OPENAI_API_KEY as environment variable.

JailGuard →

git clone https://github.com/shiningrain/JailGuard.git

Follow its README to configure MiniGPT4.

LLaVA-1.5-7B-FT → requires downloading our finetuned model and setting its path in detector_image/llava.py.

🔹 In-domain Generalization

We also provide in-domain trained versions of the Embedding-T and Embedding-I models, available in model/embedding-t/in-domain and model/embedding-i/in-domain. To use them, follow the same evaluation procedure as in the main experiments, but update the model path in detector_text/embedding-t.py and detector_image/embedding-i.py.

🏋️ Training

We provide code for training embedding-based binary classifiers for both text and image.

Text embedding classifier

python train/embedding-t.py \
  --input_dir [dir with training text jsonl files] \
  --output_dir [model output path]

JSONL format:

{"text": "example", "label": 1}   # 1 for malicious, 0 for benign

Image embedding classifier

python train/embedding-i.py \
  --input_dir [dir with training image jsonl files] \
  --output_dir [model output path]

JSONL format:

{"path": "path/to/image.png", "label": 1}

Finetuning LLaVA-1.5-7B

python train.py \
  --train_jsonl train.jsonl \
  --val_jsonl val.jsonl \
  --use_lora \
  --amp_dtype bf16 \
  --device_mode single \
  --gpu_id 0

The JSONL files should contain image paths and labels (1 = malicious, 0 = benign). Experiments in our paper use the default hyperparameters.

Tool herunterladen

Beliebteste

Alle anzeigen →

Entdecken Sie die meistgenutzten Tools unserer Community.

  • 1

    OsintHunter

    letox4/osinthunter

    🎯 Das ultimative OSINT-Toolkit: Untersuchungsgraph, Konfidenzbewertung, Alias-Familien, Karte, Zeitleiste und Berichte. 49 Module, 400+ Websites, keine API-Schlüssel erforderlich.

    867
  • 2

    ghostlock-app

    yukonga/ghostlock-app

    GhostLock One-Tap Execution App (CVE-2026-43499)

    544
  • 3

    ARTEX

    baarai/artex

    LLM-basierte Multi-Agenten-Plattform für autonome Penetrationstests mit Planer-/Worker-Agenten, Asset- und Explorationsgraphen, MITM-Verkehrsaufzeichnung und Schwachstellen-Retesting.

    385
  • 4

    redforge

    cybersecurity91/redforge

    Browser-seitiges Red-Team-Toolkit mit 102 Offline-Tools für Recon, Payload-Generierung, Credential-Angriffe, Krypto, Encoding und Engagement-Reporting.

    360
  • 5

    looking-glass

    anonghosty/looking-glass

    CSIRT-Plattform, die Shadowserver-Berichte in MongoDB einliest und sie als interaktive Karten, Analysen, Findings, Advisories und Executive-Dashboards wiedergibt.

    302
  • 6

    CyberStrikeAI

    ed1s0nz/cyberstrikeai

    Das System der Handlung für KI-native Cybersicherheit – wo Absicht zu gesteuerter Ausführung wird, Beweise zu operativem Gedächtnis werden und jeder Einsatz den nächsten verbessert.

    279

Alle Tools erkunden

Durchsuchen Sie unsere Tool-Sammlung

Alle Tools anzeigen →