Threat-model your app with your own AI agent

What is threat modeling — in plain words

Threat modeling is asking, for one area of your app: “what could an attacker do here, and are we protected?” — then writing it down as a checklist of concrete threats, each mapped to the fix and the way you’d detect it. Instead of doing that by hand, you hand a focused, expert prompt to your AI agent and it does the legwork against your actual repo. You get two complementary outputs:

How it works — 4 steps, a couple of minutes

1. 1

   Install SecureNow in your project

   In your repo’s terminal. (Optional but recommended: npx securenow login connects it to your free account so the prompt can ground its rules in your real setup.)

   $ npm install securenowCopy

   $ npx securenow loginCopy
2. 2

   Copy a category’s prompt

   Pick any of the 25 categories below and hit Copy prompt.

3. 3

   Paste it into your AI coding agent

   Open Claude Code, Cursor, Codex (or similar) at the root of your project and paste. The agent inventories your code, enumerates the threats, and audits your files — all in your environment.

4. 4

   Open the report it generates locally

   It writes four files into a new threat/<category>/ folder in your repo — two Markdown reports and two self-contained HTML pages. Just double-click the …-code-findings.html and …-detection-mitigation.html to open them in your browser — no server, no internet needed.

   threat/<category>/
   ├─ <category>-code-findings.html    # the code audit (open this)
   ├─ <category>-code-findings.md
   ├─ <category>-detection-mitigation.html # rules & defenses (open this)
   └─ <category>-detection-mitigation.md