---
title: README
emoji: 📚
colorFrom: pink
colorTo: green
sdk: static
pinned: false
---

The **Computer Incident Response Center Luxembourg (CIRCL)** is a government-driven initiative providing a coordinated and systematic response to computer security threats and incidents.

👉 [CIRCL AI Strategy](https://www.circl.lu/pub/ai-strategy/)

## Vision

Our vision is to strengthen the capabilities of our **Computer Security Incident Response Team (CSIRT)** through the strategic integration of **Artificial Intelligence**.  
AI is not intended to replace human analysts, but to **augment their expertise**, accelerate analysis, and improve decision-making.

## Mission

Our mission is to leverage AI to:

- Process, analyze, and extract value from **large and currently underutilized data sources**
- Improve operational outcomes in **threat intelligence** and **incident response**
- **Automate repetitive, well-defined, and reproducible processes**, allowing analysts to focus on higher-value tasks

This work is grounded in CIRCL’s long-standing commitment to:

- Developing **open-source tools**
- Producing **actionable threat intelligence**
- Publishing **high-quality datasets and models** for both internal use and the wider security community

## Publications

- **Bonhomme, C., & Dulaunoy, A. (2025).**  
  *VLAI: A RoBERTa-Based Model for Automated Vulnerability Severity Classification* (Version 1.4.0) [Computer software].  
  https://arxiv.org/abs/2507.03607 https://huggingface.co/papers/2507.03607

- **Bonhomme, C. (2025).**  
  *GPU Efficiency in VLAI Model Training.*  
  https://doi.org/10.13140/RG.2.2.24233.04963

## Projects


<div style="
  display: flex;
  justify-content: center;
  gap: 20px;
  flex-wrap: wrap;
  margin-top: 30px;
">

  <!-- VulnMCP -->
  <div style="
    width: 230px;
    border: 1px solid #ddd;
    padding: 20px;
    box-shadow: 2px 2px 6px rgba(0,0,0,0.08);
    background-color: white;
    border-radius: 8px;
    display: flex;
    flex-direction: column;
    text-align: center;
    min-height: 320px; /* ensures all cards same height */
  ">
    <div>
      <img src="https://raw.githubusercontent.com/vulnerability-lookup/VulnMCP/refs/heads/main/docs/VulnMCP-logo.png"
           alt="VulnMCP"
           style="display: block; margin: 0 auto 10px auto; width: 6em;">
      <p style="color: #333; font-size: 0.95em; margin-bottom: 15px;">
        A modular MCP server providing AI-driven vulnerability management skills, including severity classification and automated insights. 
      </p>
    </div>
    <a href="https://github.com/vulnerability-lookup/VulnMCP"
       rel="noreferrer" target="_blank"
       style="
         display: inline-block;
         padding: 8px 16px;
         background-color: lightcoral;
         color: white;
         text-decoration: none;
         border-radius: 5px;
         margin-top: auto;
       ">
      Learn more
    </a>
  </div>


  <!-- VulnTrain -->
  <div style="
    width: 230px;
    border: 1px solid #ddd;
    padding: 20px;
    box-shadow: 2px 2px 6px rgba(0,0,0,0.08);
    background-color: white;
    border-radius: 8px;
    display: flex;
    flex-direction: column;
    text-align: center;
    min-height: 320px; /* ensures all cards same height */
  ">
    <div>
      <img src="https://raw.githubusercontent.com/vulnerability-lookup/VulnTrain/refs/heads/main/docs/static/img/VulnTrain-logo.png"
           alt="VulnTrain"
           style="display: block; margin: 0 auto 10px auto; width: 6em;">
      <p style="color: #333; font-size: 0.95em; margin-bottom: 15px;">
        Building AI-ready datasets and training models for vulnerability management.
      </p>
    </div>
    <a href="https://github.com/vulnerability-lookup/VulnTrain"
       rel="noreferrer" target="_blank"
       style="
         display: inline-block;
         padding: 8px 16px;
         background-color: lightcoral;
         color: white;
         text-decoration: none;
         border-radius: 5px;
         margin-top: auto;
       ">
      Learn more
    </a>
  </div>

  <!-- VulnAgent -->
  <div style="
    width: 230px;
    border: 1px solid #ddd;
    padding: 20px;
    box-shadow: 2px 2px 6px rgba(0,0,0,0.08);
    background-color: white;
    border-radius: 8px;
    display: flex;
    flex-direction: column;
    text-align: center;
    min-height: 320px;
  ">
    <div>
      <img src="https://raw.githubusercontent.com/vulnerability-lookup/VulnAgent/refs/heads/main/docs/VulnAgent-logo.png"
           alt="VulnAgent"
           style="display: block; margin: 0 auto 10px auto; width: 6em;">
      <p style="color: #333; font-size: 0.95em; margin-bottom: 15px;">
        An experimental agentic cybersecurity framework combining LLMs, domain-specific tools, and distributed agents over XMPP.
      </p>
    </div>
    <a href="https://github.com/vulnerability-lookup/VulnAgent"
       rel="noreferrer" target="_blank"
       style="
         display: inline-block;
         padding: 8px 16px;
         background-color: lightcoral;
         color: white;
         text-decoration: none;
         border-radius: 5px;
         margin-top: auto;
       ">
      Learn more
    </a>
  </div>

  <!-- ML-Gateway -->
  <div style="
    width: 230px;
    border: 1px solid #ddd;
    padding: 20px;
    box-shadow: 2px 2px 6px rgba(0,0,0,0.08);
    background-color: white;
    border-radius: 8px;
    display: flex;
    flex-direction: column;
    text-align: center;
    min-height: 320px;
  ">
    <div>
      <img src="https://raw.githubusercontent.com/vulnerability-lookup/ML-Gateway/refs/heads/main/docs/Ml-Gateway-logo.png"
           alt="ML-Gateway"
           style="display: block; margin: 0 auto 10px auto; width: 6em;">
      <p style="color: #333; font-size: 0.95em; margin-bottom: 15px;">
        A FastAPI-based local inference server that loads one or more pre-trained NLP models at startup and exposes them through a clean, RESTful API.
      </p>
    </div>
    <a href="https://github.com/vulnerability-lookup/ML-Gateway"
       rel="noreferrer" target="_blank"
       style="
         display: inline-block;
         padding: 8px 16px;
         background-color: lightcoral;
         color: white;
         text-decoration: none;
         border-radius: 5px;
         margin-top: auto;
       ">
      Learn more
    </a>
  </div>

</div>


👉 [AI datasets and VLAI models](https://www.vulnerability-lookup.org/user-manual/ai/) related to the [Vulnerability-Lookup](https://github.com/vulnerability-lookup) project.