AI内容检测服务器

# Clone and run
git clone https://github.com/AdhiDevX369/MCP-S2xD.git
cd MCP-S2xD
go mod tidy
go run ./cmd/server

# Health check
curl http://localhost:3000/health

# List available tools
curl -X POST http://localhost:3000/mcp \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}'

// Request
{
  "jsonrpc": "2.0",
  "id": "req-42",
  "method": "tools/call",
  "params": {
    "tool": "analyze_origin",
    "input": {
      "text": "Furthermore, it is important to note that the implementation follows best practices.",
      "source_hint": "documentation"
    }
  }
}

// Response
{
  "jsonrpc": "2.0",
  "id": "req-42",
  "result": {
    "output": {
      "origin": "ai_system",
      "confidence": 0.85,
      "explanation": "AI indicators detected: [hedging_language, uniform_sentence_length]",
      "signals": ["hedging_language", "uniform_sentence_length"],
      "word_count": 12,
      "timestamp": "2025-11-26T09:00:00Z"
    }
  }
}

// Request - Comprehensive statistical analysis
{
  "jsonrpc": "2.0",
  "id": "stat-1",
  "method": "tools/call",
  "params": {
    "tool": "statistical_detect",
    "arguments": {
      "input": {
        "text": "It is important to note that in the evolving landscape of AI..."
      }
    }
  }
}

// Response
{
  "jsonrpc": "2.0",
  "id": "stat-1",
  "result": {
    "output": {
      "origin": "ai_system",
      "confidence": 0.78,
      "weighted_score": 0.42,
      "signals": ["ai_phrases", "uniform_structure", "zipf_deviation"],
      "entropy_result": { "average_entropy": 3.2, "origin": "ai_system" },
      "ngram_result": { "ai_phrases_found": ["it is important to note"], "origin": "ai_system" },
      "burstiness_result": { "burstiness_score": 0.18, "origin": "ai_system" },
      "zipf_result": { "r_squared": 0.72, "origin": "ai_system" }
    }
  }
}

// Request
{
  "jsonrpc": "2.0",
  "id": "ent-1",
  "method": "tools/call",
  "params": {
    "tool": "entropy_analyze",
    "arguments": {
      "input": { "text": "Your text to analyze..." }
    }
  }
}

// Response
{
  "jsonrpc": "2.0",
  "id": "ent-1",
  "result": {
    "output": {
      "origin": "human",
      "confidence": 0.72,
      "char_entropy": 4.23,
      "word_entropy": 4.56,
      "bigram_entropy": 4.89,
      "average_entropy": 4.56,
      "explanation": "High entropy indicates natural human variation."
    }
  }
}

Statistical Signals              Heuristic Signals
─────────────────────────────    ─────────────────────────────
• Low entropy (repetitive)       • Uniform sentence length
• AI-typical phrases             • Transition word density
• Uniform burstiness             • Hedging language
• Zipf's law deviation           • Structured lists
• N-gram repetition              • Explicit AI mentions

Human Indicators
─────────────────────────────
• High entropy (varied)
• Natural burstiness
• Zipf-compliant distribution
• Common typos
• Informal contractions

# Server
SERVER_PORT=:3000
SERVER_READ_TIMEOUT=10s
SERVER_WRITE_TIMEOUT=30s
SERVER_SHUTDOWN_TIMEOUT=15s

# Security
SECURITY_API_KEY_ENABLED=false
SECURITY_API_KEY=your-secret-key
SECURITY_RATE_LIMIT_RPS=100
SECURITY_MAX_REQUEST_SIZE=1048576

# Logging
LOG_LEVEL=info          # debug, info, warn, error
LOG_FORMAT=json         # json, text

# Database (optional)
DATABASE_PATH=data/mcp.db

mcp-s2xd/
├── cmd/server/          # Entry point
├── internal/
│   ├── config/          # Environment configuration
│   ├── httpserver/      # HTTP server with middleware
│   ├── logger/          # Structured logging (slog)
│   ├── mcp/             # JSON-RPC handler
│   ├── metrics/         # Request & tool metrics
│   ├── middleware/      # Security middleware stack
│   ├── rpcerror/        # Centralized error types
│   ├── storage/         # Attestation persistence
│   └── tools/           # MCP tool implementations
├── .env.example
├── go.mod
└── README.md

Request → Rate Limit → API Key Auth → Max Body Size → CORS → Security Headers → Handler

curl http://localhost:3000/metrics

{
  "uptime_seconds": 3600,
  "total_requests": 1250,
  "total_errors": 3,
  "avg_duration_ms": 2.5,
  "tool_calls": {
    "analyze_origin": 800,
    "code_origin": 300,
    "compare_texts": 150
  }
}