OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Security & Safety

Attackgen

AttackGen is an open-source Python and Streamlit application that generates cybersecurity incident-response scenarios and tabletop exercise materials using large language models, MITRE ATT&CK and ATLAS data, threat selections, and organization details.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Attackgen — OpenViST project profile in AI Security & Safety.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use AttackGen when a security team needs a structured, threat-informed starting point for an incident-response tabletop, including exercises involving AI-system attacks.

Who it is for

  • Cybersecurity and incident-response teams preparing tabletop exercises.
  • Facilitators who need scenarios tailored to an organization’s size, industry or selected threat techniques.
  • Teams developing exercises around AI/ML attacks or AI-agent insider threats.

Documented uses

  • Generate a response scenario based on a selected threat actor group or ATLAS case study.
  • Tailor an exercise to an organization’s size and industry.
  • Build custom scenarios from selected ATT&CK or ATLAS techniques or scenario templates.
  • Prepare an AI insider-threat exercise using deployment archetype, threat category, STRIDE threats and an optional scenario seed.
  • Download scenario material in Markdown and navigate key exercise sections.
Before you choose

Requirements & limits

The README describes trying the application through Streamlit Community Cloud, installing from its public GitHub repository, or pulling a Docker image. Local scenario generation requires an API key for a chosen model provider.

  • Scenario generation uses a model-provider API key; the project supports several provider APIs and custom OpenAI-compatible endpoints through LiteLLM.
  • MITRE ATT&CK and ATLAS data files are included in the repository, so the README says a fresh clone needs no separate data download.
  • The project is Python-based and lists Streamlit among its dependencies.
  • The MITRE data tools exposed by the documented MCP server are described as requiring neither an API key nor an LLM call; scenario-generation tools are a separate tier.
  • The project is licensed under GPL-3.0.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Install AttackGen from its repository

Prepare a local AttackGen installation by following the repository’s documented setup steps.

Before you start

  • Recent Python version
  • A model-provider API key for scenario generation
  • A .env file for API keys and secrets
  • MITRE ATT&CK and ATLAS data files, included in the repository

Steps

  1. Clone https://github.com/mrwadams/attackgen.git.

  2. Change into the attackgen directory.

  3. Install the required packages with pip install -r requirements.txt.

Expected outcome

The repository and its required Python packages are installed; the README lists running the app as the next stage.

Guide limitations

  • These steps cover installation only, not configuring credentials or launching the app.
  • LangSmith is optional; the README specifies a separate secrets.toml setup if using it.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
13 October 2026
Original website title
mrwadams/attackgen · Official homepage title · retained in its source language
How we review sourcesSuggest a correction