Vulnerability Database

Web Security Vulnerabilities

A practical reference covering the most critical vulnerabilities in modern web apps — with real attack examples, fixes, and how VibeWShield detects each one.

43 vulnerabilities
Critical

Cloud Storage Misconfiguration

Publicly accessible S3 buckets, GCS buckets, and Vercel/Netlify deployments expose sensitive files, environment variables, and internal build artifacts — a critical risk in AI-generated apps.

A05:2021CWE-732
Critical

Command Injection

Command injection lets attackers execute arbitrary OS commands on your server by injecting shell metacharacters into inputs that are passed to system commands.

A03:2021CWE-78
Critical

Exposed Secrets & API Keys

Hardcoded API keys, tokens, and credentials accidentally leaked in JavaScript bundles, source maps, or public endpoints — a leading cause of cloud breaches.

A02:2021CWE-798
Critical

HTTP Request Smuggling

HTTP Request Smuggling exploits discrepancies between how a frontend proxy and a backend server parse the boundaries of HTTP requests, allowing attackers to inject requests that are processed on behalf of other users.

A05:2021CWE-444
Critical

LDAP Injection

Unsanitized input in LDAP search filters enables authentication bypass, directory enumeration, and data extraction from enterprise directory services.

A03:2021CWE-90
Critical

MCP Server Security

Unauthenticated MCP servers, tool poisoning via prompt injection, and overly-permissive tool scopes give attackers direct access to AI agent capabilities — file reads, command execution, and database queries.

LLM01:2025CWE-306
Critical

Prompt Injection

Prompt injection hijacks AI/LLM-powered features by injecting instructions into user inputs, causing the model to ignore its system prompt and perform unintended actions.

LLM01:2025CWE-77
Critical

Server-Side Template Injection (SSTI)

SSTI occurs when user input is embedded into a template engine without sanitization, allowing attackers to execute arbitrary code on the server.

A03:2021CWE-94
Critical

SQL Injection

SQL Injection lets attackers manipulate database queries by injecting malicious SQL code through user input, potentially exposing, modifying, or destroying all data in a database.

A03:2021CWE-89
Critical

Web Cache Deception

Tricks CDN caches into storing personalised pages as static resources — an attacker's single crafted URL leaks the victim's private data to anyone who requests it.

A05:2021CWE-525
High

Business Logic Abuse

Business logic vulnerabilities allow attackers to exploit flaws in application workflows — bypassing payment steps, applying discounts multiple times, or escalating privileges by manipulating parameters that the application trusts without verification.

A04:2021CWE-840
High

Cross-Site Scripting (XSS)

XSS lets attackers inject malicious scripts into web pages viewed by other users, enabling session hijacking, credential theft, and full account takeover.

A03:2021CWE-79
High

DNS Misconfiguration

DNS Zone Transfer attacks, missing DNSSEC, absent CAA records, and low-TTL rebinding vectors expose infrastructure details and enable subdomain takeover in vibe-coded apps.

A05:2021CWE-16
High

GraphQL Security

GraphQL misconfigurations — enabled introspection, missing query depth limits, and unlimited batching — expose your entire API schema to attackers and enable resource exhaustion attacks that can take down your server with a single query.

A05:2021CWE-200
High

Host Header Injection

Host Header Injection tricks web applications into using an attacker-controlled domain when generating absolute URLs — poisoning password reset links, cache entries, and redirects to redirect victims to attacker infrastructure.

A03:2021CWE-20
High

HTTP/2 Attacks

H2C cleartext upgrade smuggling and CONNECT tunneling bypass proxy-layer authentication, rate limiting, and IP restrictions — reaching backend services directly.

A05:2021CWE-444
High

Insecure Deserialization

Insecure deserialization allows attackers to tamper with serialized objects to achieve remote code execution, authentication bypass, or denial of service by exploiting language-specific deserialization gadget chains.

A08:2021CWE-502
High

Insecure Direct Object Reference (IDOR)

IDOR lets attackers access other users' data by manipulating object identifiers in requests — changing user IDs, document IDs, or UUIDs to reach unauthorized resources.

A01:2021CWE-639
High

Integer Overflow & Type Juggling

AI-generated code rarely validates numeric input types. Attackers send negative prices, NaN quantities, or overflow values to manipulate business logic — free items, infinite credits, or server crashes.

A03:2021CWE-190
High

JWT Security Issues

Insecure JWT implementations — weak secrets, algorithm confusion, missing validation — let attackers forge tokens and impersonate any user including admins.

A07:2021CWE-347
High

Mass Assignment

Mass assignment vulnerabilities occur when an API binds all user-supplied fields to a data model without filtering, allowing attackers to set fields they should never control — such as role, isAdmin, or plan — and escalate their own privileges.

A01:2021CWE-915
High

NoSQL Injection

NoSQL injection manipulates MongoDB, Firebase, and other NoSQL database queries by injecting operator objects, bypassing authentication and exposing all records.

A03:2021CWE-943
High

OAuth2 Security Misconfigurations

OAuth2 misconfigurations — open redirect_uri, missing PKCE, implicit flow, exposed client secrets — allow attackers to steal authorization codes and access tokens, leading to full account takeover without knowing the victim's password.

A07:2021CWE-601
High

Path Traversal

Path traversal lets attackers access files outside the intended directory by injecting ../ sequences into file path parameters, exposing server configuration, credentials, and source code.

A01:2021CWE-22
High

Payment Security (PCI-DSS)

Vibe-coded checkout pages frequently expose raw card inputs in the DOM, skip 3DS/SCA authentication, and forget to verify Stripe webhook signatures — turning a payment form into a data exfiltration endpoint.

A02:2021CWE-319
High

Prototype Pollution

Prototype pollution lets attackers inject properties into JavaScript's Object prototype, affecting all objects in the application and potentially leading to remote code execution or authentication bypass.

A08:2021CWE-1321
High

Race Condition (TOCTOU)

Race conditions occur when multiple concurrent requests exploit a Time-of-Check to Time-of-Use window — allowing attackers to apply promo codes multiple times, double-spend balances, or claim rewards that should only be redeemable once.

A04:2021CWE-362
High

Server-Side Request Forgery (SSRF)

SSRF tricks your server into making HTTP requests to internal infrastructure — cloud metadata endpoints, internal APIs, and services that should never be reachable from the internet.

A10:2021CWE-918
High

Subdomain Takeover

Subdomain takeover lets attackers claim abandoned subdomains pointing to deprovisioned cloud services, hosting malicious content that appears to come from your trusted domain.

A05:2021CWE-350
High

Web Cache Poisoning

Web cache poisoning tricks a caching layer into storing a malicious response and serving it to all subsequent users, turning a single attacker's request into a persistent attack affecting every visitor.

A05:2021CWE-601
High

WebSocket Security

WebSocket vulnerabilities — unencrypted connections, missing authentication, and Cross-Site WebSocket Hijacking — allow attackers to intercept real-time data, send unauthorized messages, and hijack WebSocket sessions using a victim's browser cookies.

A02:2021CWE-319
High

XML External Entity (XXE)

XXE injection exploits XML parsers to read local files, perform SSRF, or cause denial of service by defining malicious external entity references in XML input.

A05:2021CWE-611
Medium

Account Enumeration

Account enumeration lets attackers discover which email addresses or usernames are registered by exploiting different server responses for valid vs. invalid accounts — enabling targeted phishing, credential stuffing, and brute-force attacks.

A07:2021CWE-204
Medium

Client-Side Path Traversal

JavaScript constructs API paths from user input without sanitization — attackers inject ../ sequences in the browser to access unauthorized endpoints, invisible to WAFs.

A01:2021CWE-22
Medium

CORS Misconfiguration

Misconfigured Cross-Origin Resource Sharing allows malicious websites to make authenticated API requests on behalf of logged-in users, stealing data and performing actions without consent.

A05:2021CWE-346
Medium

Cross-Site Request Forgery (CSRF)

CSRF tricks authenticated users into unknowingly submitting requests to your application — changing their email, transferring funds, or deleting their account — from a malicious third-party site.

A01:2021CWE-352
Medium

Dangling Markup Injection

When XSS is blocked by CSP but input is reflected without encoding, an unclosed HTML tag can exfiltrate sensitive page content to an attacker's server.

A03:2021CWE-116
Medium

Email Security (SPF / DMARC / DKIM)

Missing or misconfigured SPF, DMARC, and DKIM records allow anyone on the internet to send emails that appear to come from your domain — enabling phishing attacks on your users and damaging your brand reputation with zero technical access to your systems.

A05:2021CWE-183
Medium

HTTP Parameter Pollution

Sending the same parameter twice in a request exploits inconsistent server parsing — bypassing WAFs, overriding security parameters, and altering business logic without triggering validation.

A03:2021CWE-235
Medium

Missing Rate Limiting

Without rate limiting, attackers can brute-force passwords and OTPs, scrape your entire database, enumerate valid accounts, and spam your API with unlimited requests at zero cost.

A05:2021CWE-307
Medium

Missing Subresource Integrity (SRI)

Without SRI, a compromised CDN can serve malicious JavaScript or CSS to all your users — silently stealing credentials, injecting ads, or taking over sessions.

A08:2021CWE-353
Medium

Open Redirect

Open redirects allow attackers to craft trusted-looking URLs on your domain that redirect victims to malicious sites — enabling phishing, credential theft, and OAuth token hijacking.

A10:2021CWE-601
Low

security.txt & robots.txt Audit

Missing or misconfigured security.txt hinders responsible disclosure, while robots.txt can inadvertently reveal sensitive internal paths to attackers.

A05:2021CWE-200

Free Security Scan

Is your app vulnerable?

VibeWShield automatically tests your app for all vulnerabilities in this database using 63 scanners. Results in under 3 minutes. No signup required.

Scan your app free