Sample alert How it works Coverage PENTEST METHODOLOGY DOCS PRICING FAQ MCP CONTACT LOG IN SCAN MY APP →

Scan Your Cursor App for Vulnerabilities

ENTER YOUR CURSOR APP URL

Enter your deployed app URL to check for security vulnerabilities.

Cursor is a VS Code fork with tab completions, Composer for multi-file edits, and an Agent mode that plans, edits, and runs commands. It produces a conventional web app in whatever stack you prompted for (Next.js on Vercel, Express or Fastify, FastAPI, Rails). The attack surface is the app you deploy, not the editor.

The risk sits in the review gap. Composer can rewrite many files in one operation, and Agent mode can commit without a human reading the result. Teams accept the diff because it fixes the symptom they asked about, and miss what is gone: an auth middleware that never rejects, a parameterized query that became string concatenation, a CSRF check removed while “speeding up” a route.

Rules files add a leak path. Keys pasted into .cursorrules as a working example go to the model on every prompt and stay in git history. A black-box scan of the deployed app tests behavior, so it catches what skimming a diff misses.

Common vulnerabilities we find in Cursor apps

String-concatenated SQL

When Agent mode can’t infer your ORM - common in scripts and quick admin endpoints - it emits string-built queries that pass against test fixtures:

const rows = await db.query(`SELECT * FROM users WHERE email = '${email}'`);

The exploit is classic injection: a crafted email parameter dumps the table. Search for f"SELECT, `SELECT ${, and + user_id +; every match should become a parameterized query ($1, ?, or ORM bindings), never manual quote-escaping.

Auth middleware that never rejects

A recurring Cursor scaffold: middleware that reads the session, finds none, and calls next() anyway. The route returns 200 with {user: null, items: []}. That looks like logged-out behavior in dev, but it exposes the endpoint’s shape, admin fields included, to anyone.

Exploit: hit the API route directly with no auth header and read what comes back. Every auth middleware must end its negative path with return res.status(401).end(). Add a unit test asserting that, so a later refactor can’t silently remove it.

Broken object-level authorization (BOLA)

Generated CRUD routes authenticate the caller, then fetch by ID with no ownership check. Sign in as user B, replay user A’s request with A’s record ID, get A’s data. AI-generated frontends put real UUIDs in URLs and API calls, so collecting IDs is easy.

Add an owner_id comparison to every ID-keyed route:

if (record.owner_id !== req.user.id) return res.status(403).end();

The pattern and its variants are cataloged in BOLA in AI-generated CRUD.

Secrets in rules files and source

.cursorrules and .cursor/ files are sent to the model on every prompt, and developers paste real keys into them for context. Cursor also propagates example credentials it sees in nearby files - an .env.example value autocompletes into real source. Either way, sk-proj- or eyJ strings end up in the repo, the bundle, or a public GitHub search.

Grep .cursor/, source, and the deployed bundle for key prefixes, and check history. Removing a key without rotating it fixes nothing:

git log -p -- .cursorrules .cursor/ | grep -nE 'sk_|sk-proj-|eyJ|AKIA'

Unvalidated file uploads

Cursor’s first-pass upload handler trusts the browser: no size cap, no MIME allowlist, filename taken from the upload. That combination allows path traversal (../../ filenames), unbounded-storage DoS, and stored-payload hosting from your own domain.

Cap size, allowlist MIME by content sniff, and regenerate filenames server-side as UUIDs. The full failure catalog is in file upload, zip slip and XXE.

Permissive CORS and missing headers

cors() with no arguments or origin: '*' gets added to silence a dev-console error, then ships. Combined with credentialed requests, a hostile page can read authenticated API responses cross-origin. Generated apps also often omit CSP, HSTS, and X-Frame-Options.

Set an explicit origin allowlist and add the header baseline once in middleware (helmet covers most of it). Details in CORS and credentials misconfig.

How VibeEval works with Cursor

  1. Enter your deployed URL. Production or a preview deploy - Vercel, Netlify, Railway, anywhere. No repo access needed; the scan sees exactly what an attacker sees.
  2. The agent probes the running app. A browser-driven agent creates accounts, logs in, and works the app: replaying requests across users to find BOLA, hitting API routes directly to bypass client-side validation, inspecting bundles for leaked keys and source maps, and checking headers, CORS, and auth behavior on every endpoint it discovers.
  3. Read the report. Findings are ranked by severity with evidence, and each includes a paste-ready fix prompt. Drop the prompt into Composer or Agent mode, review the diff, redeploy, re-scan.

Manual testing vs VibeEval

Manual review VibeEval
Time per full pass Hours per Composer diff; more for Agent sessions Unattended
Cross-user (BOLA) tests Two accounts, request replay, discipline - rarely done Every scan, every ID-keyed route it finds
Regression after each AI edit Only what a reviewer happened to re-read Full re-test of the deployed app
Secrets in bundles Grep source; bundle usually forgotten Inspects what actually shipped
Business-logic flaws Strong - humans understand intent Limited - flags surface behavior
Cost Senior review time on every PR Flat, repeatable

Manual review is still the only way to judge whether the logic is right. The scanner’s advantage is repeatability: it re-runs the same attack suite after every Cursor session.

Frequently asked questions

Does VibeEval scan my source code?

No. VibeEval is black-box: it tests the deployed application the way an attacker would, with no repo access. Several issues (missing headers, leaked bundle secrets, BOLA) only show up at runtime.

What frameworks does VibeEval support for Cursor apps?

Any stack, because the scan works over HTTP. Next.js, React + Express, Vue + FastAPI, or Rails all expose the same layer where auth gaps, injection, CORS, and key leaks show up.

When should I scan during a Cursor workflow?

After any Composer or Agent session that touched auth, data access, or payments, and before promoting to production. Multi-file AI edits are when security middleware tends to regress, and a scan of the preview deploy catches it before merge.

How is Cursor different from other AI tools for security?

The generated-code patterns are shared across AI coders. Cursor’s specifics come from its workflow: Composer’s large diffs invite skimming, Agent mode can commit unreviewed code, and .cursorrules files collect pasted secrets. See Is Cursor Safe? for the IDE-level analysis.

How do I fix the vulnerabilities VibeEval finds?

Each finding ships with a remediation prompt written for an AI coding agent. Paste it into Cursor, review the diff it produces (especially what it removes), redeploy, and re-scan to confirm the finding is gone.

Test your Cursor app before launch

Point VibeEval at your deployed URL to find auth gaps, injectable queries, and leaked keys, then paste the fix prompts back into Cursor.

Scan your deployed app

Paste your live URL. We probe exposed keys, missing auth, open databases, and broken access control - results in under 60 seconds. 14-day trial.

14-day free trial · Cancel anytime

START FREE SCAN →