mcp-safeguard/index.html at main · SyedAnas01/mcp-safeguard · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<title>mcp-safeguard — MCP Security Scanner</title>
<meta name="description" content="An automated security scanner for Model Context Protocol (MCP) servers. 52 detection rules across prompt injection, credential exposure, tool poisoning, and SSRF, plus endpoint and port probes.">
<style>
* { box-sizing: border-box; margin: 0; padding: 0; }
body { font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', sans-serif; background: #0d1117; color: #e6edf3; line-height: 1.6; }
.hero { background: linear-gradient(135deg, #161b22 0%, #0d1117 100%); padding: 80px 20px; text-align: center; border-bottom: 1px solid #30363d; }
.hero h1 { font-size: 3rem; color: #f0f6fc; margin-bottom: 16px; }
.hero h1 span { color: #58a6ff; }
.hero p { font-size: 1.25rem; color: #8b949e; max-width: 600px; margin: 0 auto 32px; }
.badge-row { display: flex; gap: 8px; justify-content: center; flex-wrap: wrap; margin-bottom: 32px; }
.btn { display: inline-block; padding: 12px 24px; border-radius: 6px; text-decoration: none; font-weight: 600; font-size: 1rem; transition: all 0.2s; }
.btn-primary { background: #238636; color: white; border: 1px solid #2ea043; }
.btn-primary:hover { background: #2ea043; }
.btn-secondary { background: #21262d; color: #e6edf3; border: 1px solid #30363d; }
.btn-secondary:hover { background: #30363d; }
.install-box { background: #161b22; border: 1px solid #30363d; border-radius: 8px; padding: 16px 24px; margin: 32px auto; max-width: 500px; font-family: 'Courier New', monospace; color: #79c0ff; font-size: 1.1rem; }
.stats { display: grid; grid-template-columns: repeat(auto-fit, minmax(200px, 1fr)); gap: 24px; max-width: 900px; margin: 60px auto; padding: 0 20px; }
.stat-card { background: #161b22; border: 1px solid #30363d; border-radius: 12px; padding: 24px; text-align: center; }
.stat-number { font-size: 2.5rem; font-weight: 700; color: #58a6ff; }
.stat-label { color: #8b949e; margin-top: 8px; }
.categories { max-width: 900px; margin: 0 auto; padding: 40px 20px; }
.categories h2 { font-size: 1.8rem; margin-bottom: 32px; text-align: center; color: #f0f6fc; }
.cat-grid { display: grid; grid-template-columns: repeat(auto-fit, minmax(250px, 1fr)); gap: 20px; }
.cat-card { background: #161b22; border: 1px solid #30363d; border-radius: 12px; padding: 24px; }
.cat-card h3 { color: #f0f6fc; margin-bottom: 12px; }
.cat-card p { color: #8b949e; font-size: 0.9rem; }
.cat-badge { display: inline-block; background: #1f6feb; color: white; border-radius: 20px; padding: 2px 10px; font-size: 0.8rem; margin-bottom: 12px; }
.coverage { max-width: 900px; margin: 0 auto; padding: 40px 20px; }
.press { background: #161b22; border: 1px solid #30363d; border-radius: 12px; padding: 32px; max-width: 700px; margin: 40px auto; }
.press h2 { color: #f0f6fc; margin-bottom: 20px; }
.press-item { display: flex; align-items: center; gap: 12px; margin-bottom: 16px; color: #8b949e; }
.press-item a { color: #58a6ff; text-decoration: none; }
footer { border-top: 1px solid #30363d; padding: 32px 20px; text-align: center; color: #8b949e; }
footer a { color: #58a6ff; text-decoration: none; }
</style>
</head>
<body>

<div class="hero">
  <h1>🛡️ <span>mcp-safeguard</span></h1>
  <p>An automated security scanner for Model Context Protocol (MCP) servers. Detect vulnerabilities before deployment.</p>
  <div class="badge-row">
    <img src="https://badge.fury.io/py/mcp-safeguard.svg" alt="PyPI version">
    <img src="https://img.shields.io/badge/license-MIT-blue.svg" alt="MIT License">
    <img src="https://img.shields.io/badge/rules-52-green.svg" alt="52 Rules">
  </div>
  <div class="install-box">$ pip install mcp-safeguard</div>
  <div style="display:flex;gap:12px;justify-content:center;">
    <a href="https://github.com/SyedAnas01/mcp-safeguard" class="btn btn-primary">View on GitHub</a>
    <a href="https://pypi.org/project/mcp-safeguard/" class="btn btn-secondary">PyPI Package</a>
  </div>
</div>

<div class="stats">
  <div class="stat-card">
    <div class="stat-number">52</div>
    <div class="stat-label">Detection Rules</div>
  </div>
  <div class="stat-card">
    <div class="stat-number">4</div>
    <div class="stat-label">Attack Categories</div>
  </div>
  <div class="stat-card">
    <div class="stat-number">46%</div>
    <div class="stat-label">Servers with High/Critical Findings</div>
  </div>
  <div class="stat-card">
    <div class="stat-number">&lt;2s</div>
    <div class="stat-label">Scan Time (10K-line server)</div>
  </div>
</div>

<div class="categories">
  <h2>Detection Coverage</h2>
  <div class="cat-grid">
    <div class="cat-card">
      <div class="cat-badge">15 rules</div>
      <h3>🔴 Prompt Injection</h3>
      <p>PI-001–PI-015: Detects instructions in tool outputs that override AI behavior, role assignment attacks, and context manipulation.</p>
    </div>
    <div class="cat-card">
      <div class="cat-badge">25 patterns</div>
      <h3>🟠 Credential Exposure</h3>
      <p>Identifies API keys, OAuth tokens, AWS credentials, and other secrets exposed through tool responses.</p>
    </div>
    <div class="cat-card">
      <div class="cat-badge">40 probes</div>
      <h3>🟡 Endpoint Exposure</h3>
      <p>28 sensitive-path probes + 12 dangerous-port checks for exposed admin, debug, and metadata endpoints.</p>
    </div>
    <div class="cat-card">
      <div class="cat-badge">8 rules</div>
      <h3>🔵 Tool Poisoning</h3>
      <p>Detects malicious tool definitions that masquerade as legitimate MCP tools, including name spoofing and hidden functionality.</p>
    </div>
  </div>
</div>

<div class="press">
  <h2>📰 Coverage & Recognition</h2>
  <div class="press-item">📰 <a href="https://news.ycombinator.com/item?id=48242541">Hacker News</a> — "MCP-safeguard: Automated security scanner for MCP servers"</div>
  <div class="press-item">🔒 <a href="https://github.com/googleapis/mcp-toolbox/pull/3448">googleapis/mcp-toolbox #3448</a> — SSRF fix (CWE-918), credited to Syed Anas Mohiuddin</div>
  <div class="press-item">🔒 <a href="https://github.com/modelcontextprotocol/servers/issues/4234">modelcontextprotocol/servers #4234</a> — Security guidance filed</div>
  <div class="press-item">📊 <a href="https://github.com/OWASP/www-project-top-10-for-large-language-model-applications/issues/836">OWASP LLM Top 10 #836</a> — MCP coverage requested</div>
</div>

<footer>
  Built by <a href="https://github.com/SyedAnas01">Syed Anas Mohiuddin</a> · <a href="https://github.com/SyedAnas01/mcp-safeguard">GitHub</a> · <a href="https://pypi.org/project/mcp-safeguard/">PyPI</a> · MIT License
</footer>

</body>
</html>