AI-Powered Security CLI

Multi-agent orchestration with 34+ integrated security tools. Reconnaissance, vulnerability scanning, web testing, and automated reporting — all from your terminal.

v0.1.0 Pre-Release • Feb 17, 2026
Install with one command:
curl -sL https://raw.githubusercontent.com/dolutech/nfguard-cli/main/install.sh | sudo bash
Get Started Download v0.1.0 Read the Docs
NFGuard CLI — Interactive terminal with multi-agent AI
34+
Security Tools
4
AI Specialist Agents
3
Built-in Workflows
100%
Terminal-Based

🚧 Pre-Release — v0.1.0

This is an early pre-release version of NFGuard. It is a first step — functional, but still evolving. You may encounter bugs or rough edges as we continue to improve every aspect of the tool.

We are actively working on making NFGuard better with every update. Your feedback is invaluable to us! If you find a bug, have a suggestion, or want to request a feature, please open an issue on our GitHub repository.

NFGuard is built with love for the cybersecurity community and enthusiasts around the world. This project exists because we believe powerful security tools should be accessible to everyone. Thank you for being part of this journey. ❤️

> Core Features

Everything you need for professional security assessments, orchestrated by AI.

🤖

Multi-Agent Orchestration

An orchestrator AI delegates tasks to specialized agents (Recon, Web Testing, Vuln Scanning, Reporting) — each with its own tools and system prompt. Just describe what you need in natural language.

🛠

34+ Integrated Security Tools

From subdomain enumeration (subfinder, amass) to vulnerability scanning (nuclei, sqlmap) to web fuzzing (ffuf, feroxbuster) — all pre-bundled as compiled binaries. No manual installation needed.

💻

Interactive REPL

Rich terminal interface with tab completion, slash commands, conversation history, and real-time streaming. Chat naturally or use precise commands like /full-recon target.com.

🔒

Bash Guardrails

AI can run shell commands with your explicit approval. Built-in guardrails block dangerous patterns (rm -rf, reverse shells, privilege escalation) before you even see the prompt.

📈

Automated Reporting

Generate professional PDF and DOCX security reports directly from scan findings. The Reporting Agent structures findings by severity with executive summaries and technical remediation details.

🧰

MCP Server & Client

Run nfguard serve to expose all tools via Model Context Protocol. Also acts as an MCP client — connect external tool servers and use them from within NFGuard.

💡

Session Memory

The AI remembers targets, findings, and decisions throughout your session. Context compaction automatically summarizes long conversations when approaching token limits.

⚡

Custom Skills & Agents

Create reusable workflows as YAML skills or let the AI build them during a conversation. Define custom specialist agents with tailored tool access and system prompts.

🌐

Any LLM Provider

Works with any OpenAI-compatible API: local models (Ollama, LM Studio), OpenRouter, OpenAI, and more. We recommend running a local LLM for maximum privacy and zero API costs.

> Security Tools

34+ tools bundled as pre-compiled binaries and vendored packages. Ready to use out of the box.

Reconnaissance & OSINT

  • subfinder — Passive subdomain discovery
  • amass — Attack surface mapping (OWASP)
  • theharvester — Emails, subdomains, hosts from OSINT
  • shodan — Internet-wide device search
  • uncover — Multi-engine search (Shodan, Censys, Fofa...)
  • alterx — Subdomain wordlist permutation
  • asnmap — ASN to CIDR range mapping
  • cdncheck — CDN/WAF/cloud provider detection
  • subzy — Subdomain takeover detection
  • whois — Domain and IP registration lookup

DNS & Network

  • dnsx — Fast DNS resolution (all record types)
  • doggo — Modern DNS query with JSON output
  • naabu — Fast SYN/CONNECT port scanner
  • tlsx — TLS/SSL certificate scanner
  • mapcidr — CIDR range manipulation

Web Crawling & URLs

  • katana — Web crawler (headless browser)
  • gau — Known URLs from Wayback/CommonCrawl
  • waybackurls — Historical URLs from Wayback Machine
  • unfurl — URL component extraction
  • anew — Line deduplication utility
  • httpx — HTTP probing & tech detection
  • webfetch — In-process HTTP client (SSRF-safe)

Content Discovery & Fuzzing

  • gobuster — Directory/file brute-forcing
  • ffuf — Fast web fuzzer (FUZZ keyword)
  • feroxbuster — Recursive content discovery (Rust)

Vulnerability Scanning

  • nuclei — Template-based scanner (CVEs, misconfigs)
  • dalfox — XSS scanner (reflected, stored, blind)
  • crlfuzz — CRLF injection scanner
  • sqlmap — SQL injection detection
  • arjun — Hidden parameter discovery
  • interactsh — Out-of-band interaction (SSRF, XXE)

Reporting & Utilities

  • reportgen — PDF/DOCX report generator
  • notify — Send results to Slack/Discord/Telegram

> AI Specialist Agents

The orchestrator delegates tasks to specialized agents, each with its own tools and expertise.

🔍

ReconAgent

Network reconnaissance, DNS enumeration, subdomain discovery, OSINT gathering. Follows a least-intrusive-first approach: passive before active scanning.

🌐

WebTestingAgent

Web application security assessment: SQL injection, XSS, CRLF, parameter discovery, content fuzzing, and out-of-band vulnerability detection.

🐛

VulnScanningAgent

Cross-references services with known CVEs. Prioritizes by severity (Critical > High > Medium > Low) with CVE identifiers and remediation guidance.

📄

ReportingAgent

Generates structured PDF/DOCX security reports with executive summaries, severity-organized findings, and detailed technical remediation steps.

> Built-in Workflows

One-command automated workflows that chain multiple tools together.

/full-recon <target>

Complete reconnaissance: WHOIS lookup → DNS records (A, AAAA, MX, NS, TXT) → Port scan (top 1000 ports). Get a full picture of your target in one command.

/vuln-check <target>

Vulnerability assessment: Shodan lookup for exposed services → Nuclei scan with high/critical severity templates. Quickly identify known vulnerabilities.

/web-audit <url>

Web application audit: Nuclei full template scan → Gobuster directory enumeration (php, html, txt, bak). Comprehensive web security check.

You can also create custom skills during a conversation or define them as YAML files.

> Architecture

How NFGuard orchestrates AI agents and security tools.

┌──────────────────────┐ │ You (Terminal) │ └──────────┬───────────┘ │ ┌──────────▼───────────┐ │ Interactive REPL │ │ prompt_toolkit+Rich │ └──────────┬───────────┘ │ ┌──────────▼───────────┐ │ Orchestrator Agent │ │ LLM ↔ tool_calls │ └──┬────┬────┬────┬────┘ │ │ │ │ ┌──────────▼┐ ┌▼────▼┐ ┌▼──────────┐ ┌▼──────────┐ │ReconAgent │ │Web │ │VulnScanner │ │Reporting │ │18 tools │ │Test │ │3 tools │ │1 tool │ └───────────┘ │12 tools│ └───────────┘ └───────────┘ └───────┘ │ ┌──────────▼───────────┐ │ Security Binaries │ │ nuclei, naabu, ffuf │ │ subfinder, sqlmap... │ └──────────────────────┘