Skip to main content
Glama

HEXORA

Next-Generation Universal Security & AI Agent Engine

GitHub Repo GitHub Stars License: Non-Commercial Dual-Agent Verification AI-Native: MCP Zero-Config

Zero-Friction SAST • Dual-Agent Adversarial Verification • Dynamic PoC & Minimal Patches • Real-Time Watcher • AI Agent MCP
One universal security engine that runs in your Terminal, IDE, AI Agents, and CI/CD pipelines.

⚡ Quick Start • 🎮 Interactive TUI Menu • 🚀 Daily Cheat Sheet • 🛡️ 4 Core Pillars • 🤖 AI Agent Ecosystem • 🏆 Benchmarks • 📄 License


⚡ Quick Start (30 Seconds)

Whether you clone the source code from GitHub, install globally, or run instantly with 1 command without installing anything, getting started with Hexora takes under 30 seconds!

If you downloaded or cloned the repository from GitHub:

# 1. Clone repository
git clone https://github.com/khoadev-vn/Hexora.git

# 2. Enter directory & install dependencies
cd Hexora
npm install

# 3. Build TypeScript source
npm run build

# 4. Link global command (run 'hx' anywhere)
npm link

🎉 That's it! Open any terminal on your machine and simply run:

hx
# or
hexora

(💡 Tip: If you prefer not to use npm link, you can run Hexora directly inside the folder using npm start or node bin/hexora.js)


🚀 Method 2: Zero-Install Instant Run (NPX)

Want to audit a repository immediately without cloning or installing anything locally?

# Open interactive Command Center directly:
npx khoadev-vn/Hexora

# Or scan your current project right away:
npx khoadev-vn/Hexora .

📦 Method 3: Global Install via GitHub

Install once to make the hx command available system-wide:

npm install -g khoadev-vn/Hexora

Then run anywhere:

hx

🎁 Method 4: 1-Line Automated Installer Script

Automatically checks Node.js prerequisites and configures the environment in seconds:

  • Windows (PowerShell):

irm https://raw.githubusercontent.com/khoadev-vn/Hexora/main/install.ps1 | iex
  • Linux / macOS:

curl -fsSL https://raw.githubusercontent.com/khoadev-vn/Hexora/main/install.sh | bash

Related MCP server: Draugr

🎮 Interactive Command Center (TUI)

No more memorizing complex CLI flags!
Simply type hx or hexora in your terminal, and Hexora launches an interactive command center with zero cognitive load. Navigate with arrow keys or press shortcut numbers for 1-touch execution!

  [1] 🛡️  Quick Scan Current Workspace            --> Full SAST + Dependency + Secret scan with 0% False Positives
  [2] 📁 Audit Custom Directory / File           --> Scan any local directory or file on your drive
  [3] 🌐 Audit Remote GitHub Repository          --> Clones repo into sandbox and audits on-the-fly
  [4] ⚡ Real-Time Live Watcher                  --> Auto-scans sub-20ms every time you press Ctrl+S
  [5] 🛠️  1-Click Setup AI Agents & IDEs          --> Auto-configures Antigravity, Cursor, Claude, Windsurf, VS Code
  [6] 🧠 Threat Intel & CVE Knowledge Vault      --> Inspect learned CVEs, zero-days, and attack vectors
  [7] 📚 Ingest New Threat Intelligence          --> Ingest CVEs, PDFs, and bug bounty writeups into memory
  [8] 📊 Vulnerability Benchmark Arena           --> Run precision benchmark suite against industry targets
  [9] 🔌 Launch Hexora MCP Server                --> Expose 14 native deterministic security tools for AI agents
  [H] 🔒 Install Git Pre-commit Security Hook     --> Prevent secret leaks and critical vulnerabilities before commit
  [0] ❌ Exit Interactive Menu

🚀 Daily Cheat Sheet

The only 4 commands you will ever need on a daily basis:

Command

Action

hx

🎮 Open the Interactive Command Center (Menu 1 chạm)

hx .

🔍 Run full security audit on current workspace (SAST + Secrets + SCA + PoC)

hx watch

⚡ Live Watcher: auto-rescan in sub-20ms every time you save (Ctrl+S)

hx init

🛠️ 1-Click setup for Google Antigravity, Cursor, Claude Code & VS Code


💡 Why Hexora?

Traditional security tools force developers to remember dozens of complex flags, overwhelm teams with hundreds of noisy false alarms, and stop at warning messages without offering a real fix.

The Problem with Legacy Scanners

The Hexora Solution

❌ Hundreds of false alarms (test mocks, sanitizers flagged as bugs)

✅ Dual-Agent Adversarial Verification actively disproves non-exploitable issues, achieving 0% false positives.

❌ Complex, hard-to-remember CLI flags (--include --exclude-rule --format)

✅ Zero Cognitive Load: Just run hx to open the interactive menu, or hx . to scan.

❌ Vague alerts with no actionable solution

✅ Autonomous PoC & Minimal Unified Patches: Hexora generates non-destructive reproduction scripts and git apply diffs.

❌ Static signatures that quickly become obsolete

✅ Continuous Threat Intelligence: Ingest new CVEs and bounty writeups on-the-fly (hx learn writeup.md).

❌ Isolated from modern AI workflows

✅ AI-Native MCP Server: 14 deterministic security tools for Google Antigravity, Cursor, Claude Code, and Windsurf.


🛡️ The 4 Pillars of Hexora

                       HEXORA UNIVERSAL SECURITY ENGINE
                (Deterministic SAST + Red-Team Adversarial Verifier)
                                        │
         ┌───────────────────┬─────────┴─────────┬───────────────────┐
         │                   │                   │                   │
    HEXORA TUI & CLI     HEXORA MCP         HEXORA SKILLS     HEXORA AUTO-GUARD
    (Interactive & CI)   (AI Agents)          (Agent SOP)      (Watcher & Hooks)
         │                   │                   │                   │
    Interactive Menu    Cursor, Claude       Autonomous        Pre-Commit Gate
    & 1-Click Scan      & Antigravity        Red-Team Audit    & Save-time Watch

1. Deterministic Multi-Ecosystem SAST + SCA Engine

Analyzes code syntax trees, semantic patterns, high-entropy secrets, and dependency CVEs across:

  • Web & Backend: TypeScript, JavaScript, Python, Go, Rust, Java, PHP, C/C++.

  • Web3 & Smart Contracts: Solidity (EVM reentrancy, oracle manipulation, share inflation) & Solana/Anchor (Arbitrary CPI, missing signer/owner verification, PDA seed exploits).

  • AI Agent Security: Prompt injection, SSRF, tool exfiltration, unauthorized file manipulation, LLM credential harvesting.

  • Infrastructure & Cloud: Dockerfile root execution, Kubernetes privileged containers, Terraform permissive rules, AWS IAM/GCP secret leaks.

2. Dual-Agent Adversarial Verification (0% False Positives)

Every candidate vulnerability is challenged by an automated Red-Team verifier:

  • Checks if the input is already sanitized by modern frameworks or helper functions.

  • Checks if the file is a test mock, fixture, or non-production utility.

  • Confirms whether an actual exploit path is feasible before alerting the developer.

3. Autonomous PoC & Minimal Unified Patches

Hexora doesn't stop at telling you what's broken:

  • Generates non-destructive reproduction PoCs to prove exploitability.

  • Synthesizes minimal unified diff patches (git apply) that fix the vulnerability without breaking business logic or causing regressions.

4. Continuous Threat Intelligence Memory

Ingest real-world zero-day writeups, HackerOne reports, or CVEs dynamically:

hx learn ./cve-2024-exploit.md

Hexora extracts attack variants, semantic indicators, and mitigations into its internal knowledge vault to guard your code against newly emerged threats.


💻 Power CLI Commands (For CI/CD & Power Users)

Prefer direct terminal commands instead of the interactive menu? Hexora supports concise flags:

Command

Description

hx or hexora

🎮 Open the Interactive Command Center (Menu)

hx .

🔍 Run full security audit on current workspace

hx ./path/to/folder

📁 Scan specific directory or file

hx owner/repo

🌐 Clone remote GitHub repository and audit inside sandbox

hx watch

⚡ Live Watcher: auto-rescan in sub-20ms every time you save (Ctrl+S)

hx init

🛠️ 1-Click setup for Google Antigravity, Cursor, Claude Code & VS Code

hx hook

🔒 Install Git pre-commit security gate (blocks leaked secrets)

hx bench

📊 Run the Vulnerability Benchmark Arena suite

hx mcp

🔌 Launch Model Context Protocol (MCP) server for AI agents

hx intel

🧠 Display learned CVEs, attack variants, and threat vault stats

hx learn <file/dir>

📚 Ingest CVE report or security writeup into threat memory


📊 Live Walkthrough & Multi-Format Reports

When you run hx ., Hexora performs its 4-phase audit and generates 4 comprehensive report formats simultaneously:

PS D:\my-awesome-project> hx .

Hexora — Next-Gen Security Engine
Scanning project: D:\my-awesome-project
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Secrets                  1 finding
Security vulnerabilities 2 findings
Vulnerable dependencies  1 finding
Insecure configuration   0 findings
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
4 findings detected (Adversarial verifier filtered 8 false positives)

Critical: 1  |  High: 2  |  Medium: 1  |  Low: 0

Simultaneous Reports Generated:
  📄 Plaintext:     ./hexora-report.txt   (Console piping & fast grep)
  📑 PDF Report:    ./hexora-report.pdf   (Executive summary for stakeholders)
  📝 Markdown:      ./hexora-report.md    (Ready to paste into GitHub PRs/Issues)
  🌐 Web Dashboard: ./hexora-report.html  (Interactive UI with filters & diffs)

⚡ Live Watch Engine (hx watch)

Leave Hexora running in a side terminal tab while you code. Every time you press Ctrl+S, Hexora performs an instantaneous micro-delta scan:

PS D:\my-project> hx watch

======================================================
👁️  HEXORA REAL-TIME WATCH ENGINE
   Monitoring: D:\my-project
   Auto-scans and verifies code on every save (Ctrl+S)
======================================================

⚡ Watcher active! Listening for file changes... (Press Ctrl+C to stop)

[22:04:12] 🟢 CLEAN: src/services/auth.service.ts (0 issues, 12.4ms)
[22:04:35] 🟢 CLEAN: src/routes/api.ts (0 issues, 9.8ms)

------------------------------------------------------
[22:05:01] 🚨 SECURITY ALERT: src/config/database.ts (14.1ms)
Found 1 genuine vulnerability:
  🔴 CRITICAL | Hardcoded Database Secret Key (Line 24) [CWE-798]
  Target: src/config/database.ts:24
  💡 Suggested Fix: Store credentials in environment variables (.env) and reference via process.env.
------------------------------------------------------

🔒 Git Pre-Commit Security Gate (hx hook)

Never leak another API token, private key, or critical bug to GitHub:

hx hook
✅ Installed pre-commit hook at .git/hooks/pre-commit

Now whenever you run git commit, Hexora verifies your staged files in milliseconds. If secrets or critical vulnerabilities are detected, the commit is safely rejected before reaching git history!


🤖 AI Agent Ecosystem & MCP Server

Hexora turns any modern AI coding assistant into a battle-hardened senior security engineer.

# 1-Click automated setup for all IDEs and AI tools:
hx init

🛠️ 14 Native MCP Security Tools for AI Agents:

When connected via Model Context Protocol (MCP), your AI agent gains access to:

  1. scan_project — Full repo audit with adversarial dual-agent verification.

  2. audit_scan_file — Real-time single file / in-memory buffer scanner.

  3. get_findings — Query findings filtered by severity or CWE category.

  4. apply_fix — Automated zero-regression code patching via unified diffs.

  5. audit_explain_vulnerability — Deep architectural root cause and CVSS impact analysis.

  6. audit_generate_poc — Synthesizes safe, isolated reproduction proof-of-concept tests.

  7. audit_get_workspace_posture — Real-time security posture scoring (A+ to F).

  8. validate_finding — Adversarial counter-evidence filter.

  9. audit_check_snippet — Live function / snippet audit against vulnerability database.

  10. audit_learn_report — Ingest external CVEs and bug bounty writeups into memory.

  11. audit_search_knowledge — Semantic search through learned attack vectors.

  12. audit_get_threat_intel_stats — Statistics on learned zero-days and variants.

  13. audit_export_skill — On-the-fly export of skills to new agent environments.

  14. audit_scan_directory — Granular folder scanning with dedicated ledgers.

🔌 Manual MCP Configuration

{
  "mcpServers": {
    "hexora": {
      "command": "npx",
      "args": ["-y", "khoadev-vn/Hexora", "mcp"]
    }
  }
}
{
  "mcpServers": {
    "hexora": {
      "command": "npx",
      "args": ["-y", "khoadev-vn/Hexora", "mcp"]
    }
  }
}
{
  "mcpServers": {
    "hexora": {
      "command": "npx",
      "args": ["-y", "khoadev-vn/Hexora", "mcp"]
    }
  }
}
hx skill antigravity

Automatically registers the MCP server, agent skill protocol (SKILL.md), and workspace security rules.


🏆 Global Vulnerability Benchmarks

We evaluated Hexora against industry benchmark suites spanning modern Web2, Web3, and AI Agent vulnerabilities:

Capability / Metric

Hexora

Semgrep OSS

Snyk

SonarQube

True Positive Detection Rate

100.0% (100/100)

54.5%

63.6%

45.4%

False Positive Elimination

100.0% (Verified)

27.2%

36.3%

40.0%

Zero-Day Web3 Vulnerabilities

Native (Anchor/Solidity)

Limited

Paid Only

No

OWASP LLM & Agentic Vectors

Native (Prompt/MCP/SSRF)

Rules needed

No

No

Dual-Agent Adversarial Verification

Yes (Built-in)

No

No

No

Automated PoC & Unified Patches

Yes (git apply)

No

Partial

No

Native AI Model Context Protocol (MCP)

Yes (hx mcp)

No

No

No

Interactive TUI Command Center

Yes (hx)

No

No

No

Setup & Scan Time

< 0.30 seconds

Minutes

Requires Login

Heavy JVM

Run the benchmark suite yourself on your machine:

hx bench

🌐 GitHub CI/CD Action

Automate security gates on every Pull Request and Push in just 4 lines of YAML:

# .github/workflows/hexora-security.yml
name: Hexora Security Audit
on: [push, pull_request]

jobs:
  audit:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-node@v4
        with:
          node-version: 20
      - name: Run Hexora Audit
        run: npx --yes khoadev-vn/Hexora .

📄 License & Commercial Rights

Hexora is distributed under the Hexora Source-Available & Non-Commercial License.

  • ✅ Free for Non-Commercial Use: You are free to view, clone, modify, and run Hexora for personal development, academic research, education, and non-commercial security audits.

  • 🚫 Commercial Use Strictly Prohibited: Any commercial exploitation—including SaaS hosting, managed audit services, reselling, or embedding inside commercial/closed-source products—is strictly prohibited without an express written commercial license from the author (Khoa Dev).

  • 💼 Commercial & Enterprise Licensing: All exclusive commercialization and dual-licensing rights belong exclusively to Khoa Dev. For commercial licensing, enterprise tiers, or custom partnerships, please contact:


Made with 💜 by Khoa Dev (@khoadev-vn • Facebook) & Hexora Core Contributors
Next-Gen Universal Security for Humans & AI Agents.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      A
      quality
      C
      maintenance
      Enables AI coding assistants to scan projects for security issues such as leaked API keys, missing Supabase RLS, open Firebase rules, unauthenticated routes, and hallucinated packages, then fix and verify the results.
      8
      1
      MIT