Upload README.md

README.md

@@ -1,13 +1,87 @@
 ---
 title: Pentesting Agent Zero
-emoji: ⚡
-colorFrom: red
-colorTo: red
+emoji: 🛡️
+colorFrom: gray
+colorTo: green
 sdk: gradio
-sdk_version: 6.14.0
-python_version: '3.13'
+sdk_version: 5.0.0
 app_file: app.py
-pinned: false
+pinned: true
+mcp: true
+tags:
+  - mcp-server
+  - security
+  - pentesting
+  - agent
+  - autonomous
+  - prompt-injection-defense
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# 🛡️ Pentesting Agent Zero
+
+**Autonomous AI-Driven Penetration Testing Agent**
+
+Agent Zero is a cutting-edge autonomous penetration testing agent that combines:
+- **MCP Server Integration** — 13 pentesting tools exposed as MCP endpoints
+- **smolagents AI Agent** — Autonomous task decomposition and tool orchestration
+- **Nettacker Integration** — OWASP Nettacker for automated vulnerability scanning
+- **Automatic Prompt Injection Defense** — All model inputs automatically scanned and sanitized
+
+## 🚀 Quick Start
+
+1. Open the Space at https://huggingface.co/spaces/ScottzillaSystems/Pentesting-Agent-Zero
+2. Type a pentesting task: e.g., "Scan ports on scanme.nmap.org"
+3. Use `/help` for command reference
+
+## 🔌 MCP Integration
+
+This Space exposes all pentesting tools as MCP endpoints. Connect via:
+
+- **MCP Endpoint**: `https://scottzillasystems-pentesting-agent-zero.hf.space/gradio_api/mcp/sse`
+- **Tool Schema**: `https://scottzillasystems-pentesting-agent-zero.hf.space/gradio_api/mcp/schema`
+
+Add to your MCP client configuration:
+```json
+{
+  "mcpServers": {
+    "agent-zero": {
+      "url": "https://scottzillasystems-pentesting-agent-zero.hf.space/gradio_api/mcp/sse"
+    }
+  }
+}
+```
+
+## 🛠️ Available MCP Tools
+
+| Tool | Description |
+|------|-------------|
+| `nmap_scan` | Network port scanning with Nmap |
+| `subdomain_enum` | Subdomain enumeration via DNS & cert transparency |
+| `osint_gather` | OSINT gathering with theHarvester |
+| `web_vuln_scan` | Web vulnerability scanning (Nikto + Nuclei) |
+| `sql_injection_test` | SQL injection testing with SQLMap |
+| `password_audit` | Password/hash cracking with hashcat/john |
+| `directory_bruteforce` | Directory bruteforce with gobuster/ffuf |
+| `packet_capture` | Network packet capture with tshark |
+| `nettacker_scan` | OWASP Nettacker automated scanning |
+| `exploit_search` | ExploitDB/searchsploit lookup |
+| `generate_report` | Pentest report generation |
+| `ai_security_analysis` | AI-powered security analysis |
+| `agent_status` | Agent status and health check |
+
+## 🛡️ Prompt Injection Defense
+
+Agent Zero implements automatic prompt injection protection across all AI model interactions:
+- Pattern-based detection (18+ injection patterns)
+- Regex-based boundary detection
+- Instruction integrity seals with hash verification
+- Automatic input sanitization
+- Real-time defense logging
+
+## ⚠️ Disclaimer
+
+**FOR AUTHORIZED TESTING ONLY.** Agent Zero must only be used against systems you own or have explicit written permission to test. Unauthorized scanning may be illegal in your jurisdiction.
+
+## 📄 License
+
+MIT — See LICENSE file.