Skip to main content
Back to trending

CVEngine

AI-first external attack surface assessment with continuous discovery, evidence-based vulnerability validation, and developer-agent MCP integration.

  • ai
  • attack-surface-management
  • bbot
  • cybersecurity
  • mcp
  • nmap
  • nuclei
  • vulnerability-scanner
View on GitHub
Stars
384
Forks
57
+ today
+3
Created
3d

Ranking data as of October 26, 2026 (UTC).

Star History

Today, hour by hour

384 stars
03

Overview

AI-first external attack surface assessment with continuous discovery, evidence-based vulnerability validation, and developer-agent MCP integration. It ranks #86 on GitTiger, gaining +3 stars on October 26, 2026 (UTC).

The project is written in Python and has 57 forks. It was created 3d ago.

AI-generated summary of the repository README — may contain errors.

What it is

cvENGINE is a Python tool for assessing the external attack surface of an organisation. Its description frames it as AI-first, combining ongoing discovery with validation of vulnerabilities and an MCP interface that developer agents can call. The README's own framing is confident: it calls the tool the most advanced option of its kind available for free. The metadata lists GPL-3.0.

Capabilities described

Discovery is meant to be continuous rather than a one-off scan. Domains and their subdomains, IP addresses, open ports, detected products and versions, TLS settings and web technologies are gathered automatically into an inventory the README compares to Shodan, and that inventory is kept current. The operator supplies a domain name and the rest is handled for them.

Findings are then matched against the real environment. The README refers to a live cve.now feed plus major exploitation sources, with each item graded as fire, smoke or ice and aligned to the products and versions actually deployed, including an awareness of distribution backports. The preview also promises verdicts rather than noise, with AI involved, though that sentence is cut short in the supplied evidence, so the mechanics behind the AI step are unknown.

Getting started

Nothing in the evidence describes installation, a supported Python version, credentials for external feeds, or a first scan. Consult the repository README for those steps. The project topics point at bbot, nmap and nuclei, which hints at the scanning components, but the evidence does not explain how they are wired together.

Installation
git clone https://github.com/hevnsnt/CVEngine.git
cd CVEngine
# see README for setup

Frequently Asked Questions

What do the fire, smoke and ice ratings mean?

The README uses those three labels to grade vulnerabilities drawn from the live cve.now feed and other exploitation sources before matching them to the products and versions you run. The supplied evidence does not define the threshold behind each label.

Do I need to supply more than a domain name?

The README says you enter your domain and cvENGINE handles the rest of the discovery work. It does not state which credentials, API keys or access rights the external vulnerability feeds require.

Does cvENGINE integrate with coding agents?

Yes. The repository description cites an MCP integration aimed at developer agents, and mcp is among the project topics. The evidence does not document which tools are exposed or what parameters they take.

What language is cvENGINE written in?

The code is predominantly Python according to the repository metadata, and the license is listed as GPL-3.0. No minimum Python version or supported operating system is stated in the evidence.