Copied to clipboard!
Live Playground 4 Enterprise Pillars Execution Workflow CLI Reference Performance Benchmarks CI/CD Integrations GitHub Repository ↗

Autonomous Code Quality &
Security Intelligence Gate

Zero-dependency static security auditor engineered for high-velocity teams and AI copilots. Detect exposed credentials, backend injection vulnerabilities, React memory leaks, and invisible mobile UI traps in milliseconds.

100% Offline & Air-Gapped
< 15ms Scan Speed
Zero Dependencies
AI Prompts for Copilots
npx arnav-audit .
arnav-audit — bash — 80x24 ● node v20
$ npx arnav-audit . ▲ arnav-audit v1.1.0 — Codebase Quality Gate ℹ Architecture: In-Memory Heuristics • 100% Offline & Air-Gapped ▸ Scanning repository: /workspace/my-app ———————————————————————————————————————————————————————————————————— AUDIT REPORT OVERVIEW • Scanned 142 files in 12.4ms (91,200 lines/sec) ———————————————————————————————————————————————————————————————————— Overall Score: 100/100 (Grade A+ Enterprise Verified) Violations: 0 Critical | 0 Major | 0 Minor | 0 Suggestions 4-PILLAR CODEBASE QUALITY MATRIX: ✔ [PASS] Secret Gateways & Credentials (0 detected) ✔ [PASS] Backend Security & Weak Points (0 detected) ✔ [PASS] Memory & Resource Leakage (0 detected) ✔ [PASS] UI/UX & Accessibility Traps (0 detected) Codebase File Coverage Breakdown: • TypeScript (Backend, Core, API) 54 files (.ts) • React / Next.js Components 48 files (.tsx) • Python Services & Security Gateways 12 files (.py) • Vanilla & Tailwind Styles 18 files (.css) • JSON Manifests & Data schemas 10 files (.json) ✓ Zero internal security, leakage, or interface defects detected! All 142 files verified on this device. Meets enterprise standards. ————————————————————————————————————————————————————————————————————
Zero Install Test Drive

Live In-Browser Codebase Auditor

Experience the exact heuristic detection engine in your browser right now. Choose a test scenario or paste your own code.

Test Scenarios:
src/config/gateway.ts UTF-8 • Heuristic AST Parser
Auditor Diagnostics HUD Scan: 1.2ms

Comprehensive 4-Pillar Codebase Inspection

Grounded in concrete AST heuristics, pattern boundary recognition, and mobile viewport constraints.

PILLAR 01 • CWE-798

Secret Gateways & Credentials

Flags exposed high-privilege credentials committed to code trees, unencrypted RSA private keys, plain-text database URIs, and leaked third-party service tokens before git commits.

AWS Keys (AKIA...) OpenAI & Claude Keys GitHub PAT (ghp_...) Stripe (sk_live_...) Postgres / Redis URI Slack & Discord Hooks Private RSA / PEM Keys
PILLAR 02 • CWE-89 & 78

Backend Security & Weak Points

Detects SQL string interpolation, shell command injection vectors, SSRF requests without IP boundaries, unsafe pickle deserialization, wildcard CORS, and debug modes left enabled in production.

Raw SQL String Formatting shell=True / exec() SSRF Loopback Vectors verify=False TLS FastAPI DEBUG=True Insecure Deserialization Wildcard CORS Credentials
PILLAR 03 • CWE-400

Memory & Resource Leakage

Catches dangling browser event listeners in React lifecycle hooks without teardown, unmount timer leaks, Python mutable default arguments, and state pollution of global window instances.

useEffect removeEventListener setInterval Teardown Python def foo(items=[]) Global window Pollution setMaxListeners Silencing Residual console.log
PILLAR 04 • WCAG 2.4.7

Invisible UI/UX Traps & A11y

Flags subtle mobile interface defects: iOS Safari auto-zooming on inputs, 300ms double-tap latency delays, squished SVG flex icons, 100% viewport scrollbar traps, and destroyed keyboard focus outlines.

iOS <16px input zoom touch-action: manipulation flex-shrink: 0 on SVGs outline: none (WCAG 2.4.7) 100vw Viewport Bleed Missing image alt tags
Zero-Friction Workflow

Installation to Final Execution Pipeline

A clear 4-step workflow: from initial setup to scanning and automated AI remediation.

01

Install Globally

Install arnav-audit globally via npm or run instantly without install using npx.

npm i -g arnav-audit
Or GitHub: npm i -g github:arnnnnaaavvvvv/MEJOR
02

Verify CLI Binary

Confirm the global executable is correctly registered and ready to audit.

arnav-audit --version
Active version: arnav-audit v1.1.0
03

Run Codebase Audit

Scan your project repository for security leaks, memory issues, and UI defects.

arnav-audit .
Specific folder: arnav-audit ./src
04

Compile AI Fix Prompts

Export verified defect locations and instructions directly for Cursor, Claude Code, and Antigravity.

arnav-audit . --prompts
Raw JSON: arnav-audit . --json
Extreme Engineering

Speed & Footprint Benchmarks

Engineered in pure native JavaScript with zero external runtime dependencies. Runs hundreds of times faster than heavy legacy scanners.

Execution Latency
< 15ms

Instantaneous AST-heuristic scanning in pure local memory. Zero JVM or heavy language daemon warmup required.

arnav-audit 12.4ms
Legacy SAST 42,000ms
Runtime Dependencies
0 Deps

Zero sub-dependencies in node_modules. Completely immune to supply-chain attacks, left-pad vulnerabilities, and nested package bloat.

arnav-audit 0 packages (2.8MB)
Heavy Linters 1,280+ packages (450MB)
Network & Air-Gap
0 Bytes

Zero external network requests or telemetry emitted. Proprietary enterprise source code never leaves your workstation or CI runner.

arnav-audit 100% Offline
Cloud SaaS Scanners Uploads entire repo
Command Line Interface

Command-Line Reference

Full set of flags supported by the zero-dependency CLI binary.

Command / Flag Alias Category Description Quick Copy
arnav-audit [target] — Audit Scan target directory or single file (defaults to current directory .)
--prompts --fix Output Generate pre-compiled, copy-paste AI fix prompts for Cursor, Claude Code & Antigravity
--files --list Audit Output every single verified file path with individual pass/fail inspection status
--security-only --security Pillar Filters Scan exclusively for exposed credentials, secret gateways & injection vectors
--backend-only --backend Pillar Filters Scan exclusively for backend weak points, SQLi, RCE, SSRF & debug configurations
--leakage-only --leakage Pillar Filters Scan exclusively for event listener, timer, window pollution & memory leaks
--ui-only --ui Pillar Filters Scan exclusively for invisible UI/UX mobile traps, iOS zoom & accessibility defects
--json — Output Output machine-readable JSON object for automated CI/CD pipeline gating
--ci — CI/Output Run in strict CI mode with zero-tolerance exit codes on critical violations
--verbose — Audit Display detailed scan logs, extensions breakdown and file-by-file verification trace
-v, --version — Audit Display installed package version (e.g. arnav-audit v1.1.0)
Enterprise Automation

Automated CI/CD & Copilot Workflows

Enforce codebase quality automatically across GitHub pull requests, Cursor, Claude Code, and Husky.

GitHub Actions Quality Gate

Block pull requests containing security leaks or mobile traps automatically with exit status code checks.

name: Codebase Quality Gate
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 arnav-audit Quality Gate
        run: npx arnav-audit . --ci
Showcase Your Quality

Interactive README Badge Generator

Add verified security and quality gate status to your GitHub repository README with one click.

Live Badge Preview:
Quality Gate Badge
[![arnav-audit](https://img.shields.io/badge/quality%20gate-Grade%20A%2B%20%E2%9C%93-6366f1?style=flat-square&logo=shield)](https://mejor-iota.vercel.app/)

100% Offline & Air-Gapped Security Guarantee

arnav-audit performs all inspections strictly in local process memory using zero external runtime dependencies. No code is transmitted to cloud servers, no telemetry is gathered, and no network connections are opened during scans. Completely safe for proprietary enterprise repositories, air-gapped CI runners, and defense-grade environments.

Zero Telemetry
Zero Dependencies
Air-Gapped Ready
Clear Answers

Frequently Asked Questions

Everything you need to know about arnav-audit architecture and engineering standards.

How does arnav-audit achieve zero external dependencies?
arnav-audit is written directly against the Node.js standard runtime APIs (fs, path, perf_hooks). By leveraging high-performance streaming regex engines and targeted AST heuristics, it eliminates the need for massive third-party parser trees, making it instantly executable on any system with Node >= 16.
Is my source code or proprietary IP ever transmitted anywhere?
Never. arnav-audit makes zero network requests, has no telemetry collection, and initiates zero external socket connections. All file reads and heuristic validations occur 100% within your local device's memory. It is fully approved for air-gapped defense networks and private corporate repositories.
How does the AI prompt generation (--prompts) work?
When run with --prompts (or --fix), arnav-audit groups detected violations by file path and line number, pairs each with its corresponding CWE/OWASP remediation rule, and formats the output into a structured markdown prompt optimized for LLMs like Cursor Composer, Claude Code, and Antigravity. You simply copy and paste the prompt into your AI agent to auto-fix the entire codebase.
Can arnav-audit be used as a pre-commit hook?
Yes! Because arnav-audit executes in under 15ms, it introduces zero noticeable delay during git commits. You can configure it inside Husky or standard .git/hooks/pre-commit to block developers from accidentally committing credentials, unmounted React listeners, or broken mobile viewport traps.
How does it detect subtle mobile UI and iOS auto-zoom traps?
Pillar 4 analyzes CSS, SCSS, and HTML styles against mobile platform constraints. For instance, iOS Safari forces a viewport zoom whenever an input font-size is smaller than 16px, breaking mobile layouts. arnav-audit flags these declarations, as well as missing touch-action: manipulation and destroyed keyboard :focus-visible indicators (WCAG 2.4.7).