SKIP TO CONTENT

Secure My App

  • Plugin
  • Skill

Find what a stranger can take from your app.

  • Fixes proven closed

    A fix counts only when its VERIFY output changes. All 11 sample findings have one.

  • Proof for each hole

    Each finding shows its proof. make test replays all 11 sample findings.

  • No surprise changes

    The skill writes a report and stops. Each fix waits for your yes.

Works with: Any agent that reads skills. Full runs tested: Claude Code, Codex CLI. Not tied to one model: your agent's model follows a Markdown skill.

Pick your agent

npx skills add Aboudjem/secure-my-app -g

Works in more than 70 agents: pick yours in the menu. Supported agents (new tab)Installs for all your projects. Only this project: drop -g.Its page on skills.sh (new tab)

Limit: Secure My App is not a penetration test (a paid attack by an expert).

github.com/Aboudjem/secure-my-app (new tab)Version 0.3.2

What it does

  • Find holes before attackers do. It runs 65 named checks in seven groups.
  • Trust each finding. PROOF shows the code line and, with a live URL, the real request and its output.
  • Check the fix yourself. Each finding has a VERIFY command whose output must change.
  • Know what is not proven. What it cannot settle is marked UNPROVEN, never passed.
  • Stay in charge: it writes a report and stops. Nothing changes until you approve a fix.

How it works

  1. 1. It reads the app and runs the checks in the order an attacker tries them.
  2. 2. It writes .secure-my-app/report.md. The report opens with the verdict and the three things to do first.
  3. 3. You approve one fix. Nothing in your repository changes before that.
  4. 4. It runs the same command again to prove that the hole is shut.

Example

One finding from the README, abridged. It comes from a small test app with holes left in on purpose.

HIGH R4 Any signed-in user can read any invoice by changing the number in the URL
WHERE   server.js:83-92 (the lookup is line 89)
PROOF   [static]   server.js:89   const invoice = invoices.find((i) => i.id === id);
        [executed] alice (team_1) reads bob's invoice:
                   GET /api/invoices/1042 -> 200 {"id":1042,"owner":"u_2","team":"team_2",...}
FIX     const invoice = invoices.find((i) => i.id === id && i.owner === user.id);
VERIFY  alice:  GET /api/invoices/1042 (bob's)   -> 404 "no such invoice"
        alice:  GET /api/invoices/1041 (hers)    -> 200, the invoice
STATUS  OPEN

More ways to install

Claude Code plugin

claude plugin marketplace add Aboudjem/10x
claude plugin install secure-my-app@10x

Agent skill, for Cursor, Codex and other agents

npx skills add Aboudjem/secure-my-app

Limits

  • A clean run is not proof that your app is secure. The third gate label reads NOTHING CONFIRMED HERE.
  • It is not a penetration test, and nobody signs the report.
  • Without a live URL, most access checks come back UNPROVEN.
  • The only published run is on a small test app, not on a production app.
Good for
A review before launch, where you want to see the proof of each hole.
Not for
A signed security audit or a penetration test.

Questions

What is Secure My App?

Secure My App is an open-source agent skill that finds what a stranger can take from your app. make test replays the proof behind all 11 sample findings.

How do I install Secure My App in Claude Code?

Run claude plugin marketplace add Aboudjem/10x, then claude plugin install secure-my-app@10x. Then say secure my app before I launch.

Which agents does Secure My App work with?

Tested on 17 November 2026: full runs in Claude Code and Codex CLI. Other agents that read skills are documented, not tested.

Is Secure My App free?

Yes, it is open source under the MIT license. It also installs in other agents with npx skills add.