Metadata-Version: 2.4
Name: nexargio-mcp
Version: 1.1.1
Summary: Nexargio Sovereign Cyber-Physics Threat Scanner for MCP
Author-email: Nexargio Technologies <engineering@nexargio.com>
License: Proprietary
Project-URL: Homepage, https://nexargio-vip-gateway-bepltxkx4a-el.a.run.app
Project-URL: Documentation, https://nexargio-vip-gateway-bepltxkx4a-el.a.run.app/portal/api-keys
Keywords: mcp,model-context-protocol,cybersecurity,zero-day,phishing-detection,threat-intelligence,quishing
Classifier: Development Status :: 5 - Production/Stable
Classifier: Intended Audience :: Developers
Classifier: Intended Audience :: Information Technology
Classifier: Topic :: Security
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Requires-Python: >=3.10
Description-Content-Type: text/markdown
Requires-Dist: mcp<2.0.0,>=1.0.0
Requires-Dist: httpx<1.0.0,>=0.27.0
Requires-Dist: numpy>=1.24.0
Requires-Dist: opencv-python-headless>=4.8.0
Provides-Extra: test
Requires-Dist: pytest>=7.4.0; extra == "test"
Requires-Dist: qrcode>=7.4.0; extra == "test"

# Nexargio MCP Server
Real-time phishing and web-threat analysis for AI agents.

Connects **Claude Desktop**, **Cursor IDE** and other MCP-compatible agent frameworks to the Nexargio threat-analysis service.

---

## The tool: `nexargio_scan`

One tool, one call:

```
nexargio_scan(target: str, scan_mode: str = "fast")
```

### What `target` accepts
1. **Full URL:** e.g. `https://login.microsoftonline.portal-security-check.com/auth`
2. **Domain name:** e.g. `portal-security-check.com` (normalized to `https://`)
3. **QR code image:** a local image path or Base64 string (the QR code is decoded and its destination is scanned)

### What you get back
- **Verdict, threat score (0-100) and recommended action** (allow, isolate, or block)
- **Phishing and credential-theft detection**
- **Infrastructure intelligence** (domain age, TLS certificate, hosting and routing)
- **Brand impersonation detection** (lookalike domains and mimicry of major brands)
- **Cloaking and redirect awareness** (the final destination and redirect chain)
- **AI analysis** with a plain-language explanation of *why* a URL was judged the way it was, plus the supporting evidence
- **MITRE ATT&CK mapping** for the techniques observed
- **Tamper-evident evidence seal** (SHA-256) for audit trails

If a scan cannot be completed, the result is `SCAN_INCOMPLETE` with the action `ERROR_REVIEW_REQUIRED`. Treat that as "not verified", never as safe.

---

## Get an API key

Create a key on the **API Keys** page of the [Nexargio portal](https://nexargio-vip-gateway-bepltxkx4a-el.a.run.app/portal/api-keys). The full key is shown once, so store it securely.

---

## Connect to Claude Desktop

In `%APPDATA%\Claude\claude_desktop_config.json` (macOS: `~/Library/Application Support/Claude/claude_desktop_config.json`):

```json
{
  "mcpServers": {
    "nexargio": {
      "command": "uvx",
      "args": ["nexargio-mcp"],
      "env": {
        "NEXARGIO_API_KEY": "<YOUR_API_KEY>"
      }
    }
  }
}
```

## Connect to Cursor IDE

In `.cursor/mcp.json`:

```json
{
  "mcpServers": {
    "nexargio": {
      "command": "uvx",
      "args": ["nexargio-mcp"],
      "env": {
        "NEXARGIO_API_KEY": "<YOUR_API_KEY>"
      }
    }
  }
}
```

---

## Verify it works

Restart your client, then ask your agent:

> Scan https://example.com with nexargio_scan.

You should get a structured result with a verdict, threat score and evidence seal.
