DEV Community: Rudra Sarker

🔬 3D Science Lab — Interactive 3D STEM Education with 40+ Experiments Built Using Next.js and Three.js

Rudra Sarker — Wed, 01 Apr 2026 17:33:28 +0000

Making Science Interactive

Traditional science education relies on static textbook diagrams and 2D illustrations. But science happens in three dimensions. I built 3D Science Lab to make STEM education immersive — allowing students to interact with experiments in 3D, rotate models, zoom in on details, and truly understand the science behind what they see.

What is 3D Science Lab?

3D Science Lab is an interactive web platform featuring 40+ 3D science experiments across four core disciplines:

Physics — mechanics, optics, waves, electricity
Chemistry — molecular structures, reactions, periodic table in 3D
Biology — cell structures, organ systems, DNA
Mathematics — geometric shapes, functions, calculus visualizations

Key Features

🧪 40+ Interactive Experiments

Each experiment is fully interactive — drag, rotate, zoom, and manipulate to explore scientific concepts hands-on.

🎮 Immersive 3D Rendering

Built with Three.js and React Three Fiber, the platform delivers smooth, WebGL-powered 3D graphics directly in the browser. No downloads, no plugins.

📱 Responsive Design

Works on desktop, tablet, and mobile. Science class shouldn't require a specific device.

⚡ Fast Performance

Optimized rendering pipeline ensures smooth 60fps interactions even with complex 3D models.

Tech Stack

Framework: Next.js 15
3D Engine: Three.js + React Three Fiber
Language: TypeScript
Animation: Framer Motion
UI Controls: Leva
Post-processing: React Three Postprocessing

Live Demo

🔗 3D Science Lab — Explore experiments now

Why It Matters

Studies show that interactive 3D learning improves retention by up to 80% compared to traditional 2D methods. 3D Science Lab brings this capability to every student with a browser — no expensive lab equipment needed.

*Built by Rudra Sarker — Open Source Developer

Connect: X/Twitter | LinkedIn | GitHub

🥷 StealthHumanizer — A Free Open-Source AI Text Humanizer with 13 Providers and Multi-Pass Ninja Mode

Rudra Sarker — Wed, 01 Apr 2026 17:32:45 +0000

Why StealthHumanizer?

With the rise of AI-generated content, tools that can humanize text are in high demand. But most solutions are paid, require sign-ups, or limit your usage. I wanted to build something different — a completely free, open-source text humanizer that anyone can use without restrictions.

StealthHumanizer supports 13 text generation providers, 4 rewrite levels, 13 distinct tones, and a multi-pass "ninja mode" for maximum naturalness.

Features

🔄 13 AI Providers

StealthHumanizer works with OpenAI, Anthropic, Google, Mistral, Cohere, and many more providers. Switch between them freely — whatever works best for your content.

📊 4 Rewrite Levels

From light touch-ups to complete rewrites, choose the level that fits your needs.

🎨 13 Tones

Professional, casual, academic, creative, persuasive, and more. Pick the tone that matches your content's purpose.

🥷 Multi-Pass Ninja Mode

Run multiple humanization passes for the most natural-sounding output. The ninja mode applies layered transformations to make text virtually indistinguishable from human writing.

🔓 No Login, No Limits

Zero friction. No account creation. No usage caps. Just open it and use it.

Tech Stack

Language: TypeScript
Architecture: Provider-agnostic plugin system
UI: Clean, minimal interface focused on usability

Live Demo

🔗 GitHub Repository

Why Open Source?

Text humanization should be accessible to everyone — students, content creators, developers, and researchers. Locking this behind paywalls creates an uneven playing field. StealthHumanizer levels it.

*Built by Rudra Sarker — Open Source Developer

Connect: X/Twitter | LinkedIn | GitHub

🧠 MindWell — A Free, Open-Source Mental Health Platform Built with React and Next.js

Rudra Sarker — Wed, 01 Apr 2026 17:32:11 +0000

Why I Built MindWell

Mental health is one of the most critical challenges of our generation, yet access to quality mental health resources remains limited — especially in developing countries. As a student from Bangladesh, I wanted to build something that makes mental health support accessible to everyone, regardless of their location or financial situation.

MindWell is a free, open-source mental health education and support platform built with a safety-first approach.

Key Features

🤝 Peer Support Community

MindWell includes a built-in peer support system where users can share their experiences, offer encouragement, and connect with others who understand what they are going through. Community is at the heart of recovery.

🚨 Crisis Resources

Instant access to crisis helplines and emergency resources. Users can find help immediately when they need it most — no digging through websites, no confusion.

♿ Accessibility-First Design

Every component was designed with accessibility in mind. Screen reader support, keyboard navigation, high contrast mode, and semantic HTML ensure that MindWell works for everyone.

🔒 Privacy & Safety

No data collection beyond what is necessary. No ads. No tracking. Mental health is sensitive — the platform respects that.

Tech Stack

Frontend: React 19 + Next.js 15
Styling: Tailwind CSS
Deployment: Vercel
Design: Component-driven architecture with reusable UI patterns

Why Open Source?

Mental health tools should not be locked behind paywalls. By making MindWell open source, other developers can contribute features, translate the platform into their local languages, and adapt it for their communities.

Live Demo

🔗 MindWell — Try it yourself

GitHub

🔗 rudra496/mindwell — Star it, fork it, contribute

What Makes MindWell Different?

Unlike generic wellness apps, MindWell was designed from the ground up as a mental health-first platform — not a fitness tracker that added a meditation feature. Every design decision prioritizes the user's emotional safety and accessibility.

If you or someone you know could benefit from this, please share it. Mental health matters, and open source can make a real difference.

Built by Rudra Sarker — AI Developer & Open Source Contributor

Connect with me: X/Twitter | LinkedIn | GitHub

🤖 TermMind — A Free, Open-Source AI Terminal Assistant That Rivals Claude Code and Aider

Rudra Sarker — Fri, 27 Mar 2026 11:18:48 +0000

An AI-powered terminal assistant with 7 providers, 40+ commands, code analysis, file editing with undo, snippet management, and project templates — all in pure Python with zero cost options.

The Problem

AI coding assistants like Claude Code and Aider are incredible — but they're either:

Expensive (Claude Code: $20-200/month)
Locked to one provider (Claude Code → Anthropic only)
Heavy (Node.js, lots of dependencies)
No free tier for local/offline use

I wanted something that:

Works with any AI provider
Has a free tier (Gemini, Groq, Ollama)
Runs in the terminal with a beautiful UI
Is truly open source

What Is TermMind?

TermMind is a modern, open-source AI terminal assistant. Think Claude Code or Aider — but free, open-source, and lightweight.

pip install termmind
termmind init    # Select provider & API key
termmind chat    # Start chatting with AI

Supported Providers (7!)

Provider	Free Tier?	Setup
Google Gemini	✅ Yes (free API key)	aistudio.google.com
Groq	✅ Yes (free tier)	console.groq.com
Ollama	✅ Yes (fully offline!)	`ollama pull llama3.2`
OpenAI	❌ Paid	platform.openai.com
Anthropic Claude	❌ Paid	console.anthropic.com
Together.ai	❌ Paid	together.ai
OpenRouter	❌ Paid	openrouter.ai

Yes, you can use TermMind completely free with Gemini, Groq, or Ollama.

Core Features

🎨 Beautiful Terminal UI

Syntax highlighting
Markdown rendering
Streaming responses
Dark/Light themes

✏️ Smart File Editing

AI-powered file creation and editing
Side-by-side diff view in the terminal
Hunk-by-hunk confirmation (accept/reject individual changes)
Undo system — every edit is tracked

📁 Smart Context Engine

Automatically includes relevant files from your project
.termmindignore support (like .gitignore)
Persistent code index — function signatures, class definitions, imports
Supports 9 languages: Python, JS, TS, Go, Rust, Java, Ruby, C, C++

🐙 Git Integration

Status, diff, log
AI-generated commit messages

📊 Cost Tracking

Real-time token count and cost estimation

💾 Session Management

Save and load conversations

🔄 Provider Switching

Change models mid-conversation with /provider

40+ Commands

CLI Commands

termmind init        # Configure provider & API key
termmind chat        # Interactive chat session
termmind ask "q"     # One-shot question
termmind edit file   # Edit a file with AI
termmind review ./src  # Review code in directory
termmind explain file  # Explain a file
termmind test file     # Generate tests
termmind history       # Show saved sessions
termmind config        # Show current configuration

In-Chat Slash Commands

/edit <file> [instruction]  — Edit a file with AI
/run <command>              — Run a shell command
/files                      — List files in context
/search <query>             — Search project files
/tree                       — Show file tree
/save [name] / /load <name> — Session management
/model / /provider          — Switch mid-conversation
/cost                       — Show token usage & cost
/undo                       — Undo last file edit
/diff                       — Show session changes
/snippet save/search/load   — Code snippet management
/template list/use          — Project scaffolding
/refactor <op> <file>       — AI-powered refactoring

Standout Features

📦 Snippet Manager

Save code snippets from conversations
Search by name, description, tags, or code content
Template variables: {{filename}}, {{datetime}}, {{year}}, etc.
Auto-suggest relevant snippets based on context

🏗️ Project Templates

Scaffold projects from 8 built-in templates:

Python, FastAPI, React, Next.js, and more

🔧 Refactoring Engine

AI-powered refactoring with:

Diff preview
Confirmation prompt
Full undo support

🐳 Docker Support

# Build and run
docker compose up --build

# Fully offline with Ollama
docker compose --profile local-llm up --build

Shell Completions

Auto-generated completions for Bash, Zsh, and Fish:

termmind completions install

How It Compares

Feature	TermMind	Claude Code	Aider
Open Source	✅ MIT	❌	✅ Apache
Free Tier	✅ (Gemini/Groq)	❌	❌
Local/Offline	✅ (Ollama)	❌	✅
Multi-Provider	✅ 7 providers	❌	✅
Rich Terminal UI	✅	✅	❌
Undo Edits	✅	❌	✅
Cost Tracking	✅	❌	❌
Session Save/Load	✅	❌	❌
Dependencies	4 packages	Heavy	Moderate

The Architecture

CLI (click commands)
├── Commands Layer (chat, ask, edit, review…)
├── Diff Engine (side-by-side diffs)
├── Context Engine (smart file selection, code index)
│   ├── File Ops (read, write, edit)
│   ├── Git Module (status, diff, log)
│   └── API Client (streaming, multi-provider)
└── Plugin System (on_start, on_msg hooks)

Get Started

pip install termmind
termmind init      # Walks you through setup
termmind chat      # Start chatting!

Or for free usage:

# Get a free Gemini API key from aistudio.google.com
termmind init      # Select "gemini" as provider
termmind chat      # Start chatting — it's free!

⭐ GitHub: github.com/rudra496/termmind
📦 PyPI: pypi.org/project/termmind

Rudra Sarker — student researcher and full-stack developer from Bangladesh. Building free, open-source developer tools because great tools shouldn't cost a fortune.

🔗 Connect:

GitHub | LinkedIn | Twitter/X | ResearchGate | Personal Site

🗺️ DevRoadmaps — A Free, Open-Source Roadmap.sh Alternative with 17 Career Paths and 1,700+ Resources

Rudra Sarker — Fri, 27 Mar 2026 11:12:32 +0000

17 comprehensive developer roadmaps, 800+ topics, 1,700+ free resources, offline support, and community features — all in pure HTML/CSS/JS with zero dependencies.

The Problem

roadmap.sh is great — but it has limitations:

No interactive progress tracking
Limited resources
Print/poster is paid
No community features
No offline support
No resource filtering or ratings

I wanted something that goes further — truly free, feature-rich, and community-driven.

What Is DevRoadmaps?

DevRoadmaps is a free, open-source developer roadmap platform with 17 career paths, 800+ curated topics, and 1,700+ free resources.

Live Demo

🌐 rudra496.github.io/devroadmaps

17 Complete Roadmaps

#	Roadmap	Topics	Description
1	🎨 Frontend Developer	52	HTML, CSS, JS, React, Vue, Angular
2	⚙️ Backend Developer	50	Python, Node, Go, APIs, databases, Docker
3	🌐 Full Stack Developer	50	End-to-end web development
4	🧠 ML / AI Engineer	50	ML, deep learning, LLMs
5	☁️ DevOps / Cloud	50	Docker, K8s, AWS, Terraform, CI/CD
6	📱 Mobile Developer	50	Flutter, React Native, Swift, Kotlin
7	🔒 Cybersecurity	50	Pentesting, OWASP, cloud security
8	🗃️ Data Engineer	50	Spark, Kafka, Snowflake, Airflow
9	⛓️ Blockchain / Web3	50	Solidity, DeFi, smart contracts
10	🎮 Game Developer	50	Unity, Godot, Unreal, C#, C++
11	🔌 Embedded / IoT	50	Arduino, ESP32, RTOS, sensors
12	🛡️ DevSecOps Engineer	40	Container security, SAST/DAST, zero trust
13	🧪 QA / Test Engineer	43	Selenium, Playwright, BDD, TDD
14	✍️ Technical Writer	42	API docs, Docusaurus, Mermaid
15	⚡ Low-Code / No-Code	32	Bubble, Webflow, Zapier, Airtable
16	🏗️ Cloud Architect	42	AWS, GCP, Azure, IaC, FinOps

Plus pre-built learning paths like:

Full Stack in 6 Months (~22 topics)
Zero to ML Engineer (~12 topics)
DevOps Career Switch (~14 topics)

Features That Set It Apart

✅ Interactive Progress Tracking

Track your learning with clickable nodes. See your progress visually.

💬 Community Tips

Share and read tips for each topic. Learn from others' experiences.

⭐ Resource Ratings

Rate resources 1-5 stars. Find the best learning materials faster.

🔍 Resource Filtering

Filter by type: 📖 Docs, 🎥 Videos, 🎯 Courses, 🛠️ Tutorials, 🔧 Tools

📥 Export & Bookmarks

Export community tips as JSON
Bookmark topics across roadmaps
Export bookmarks as a study plan (Markdown)

🍅 Built-in Pomodoro Timer

Study timer with break reminders integrated into each roadmap.

🏆 Achievement System

Unlock badges as you learn:

🌟 First Steps (complete your first node)
🔥 Getting Serious (complete 10 nodes)
🏆 Roadmap Master (complete an entire roadmap)
🧠 Polymath (start 5 different roadmaps)

🌙 Dark/Light/Auto Theme

Three theme modes with smooth transitions. Your preference is persisted.

📱 PWA Support

Works offline as a Progressive Web App. Install it on your phone or desktop.

🖨️ Print as Poster or Checklist

Print any roadmap as a poster or a progress checklist.

How It's Built

HTML5 — Semantic markup
CSS3 — Custom properties, glassmorphism, animations
Vanilla JavaScript — Zero dependencies
GitHub Pages — Hosting
MIT License — Free forever

Just clone and open index.html — no build step needed:

git clone https://github.com/rudra496/devroadmaps.git
cd devroadmaps
python3 -m http.server 8000
# Open http://localhost:8000

Comparison with roadmap.sh

Feature	DevRoadmaps	roadmap.sh
Progress tracking	✅	❌
800+ curated topics	✅	~300
1,700+ free resources	✅	Limited
Community tips	✅	❌
Resource ratings	✅	❌
Resource filtering	✅	❌
Search within roadmaps	✅	❌
Print poster	✅ (free)	✅ (paid)
Offline PWA	✅	❌
100% free	✅	Partial

Who Is This For?

Self-taught developers who need structured learning paths
Students planning their career trajectory
Career switchers looking for clear roadmaps
Teams who want to standardize onboarding

Get Started

🌐 Live: rudra496.github.io/devroadmaps
⭐ GitHub: github.com/rudra496/devroadmaps

Rudra Sarker — student researcher and full-stack developer from Bangladesh. Building free, open-source tools for developers worldwide.

🔗 Connect:

GitHub | LinkedIn | Twitter/X | ResearchGate | Personal Site

🔍 CodeVista — A Code Analysis Tool That Generates Beautiful HTML Reports with Zero Dependencies

Rudra Sarker — Fri, 27 Mar 2026 10:54:42 +0000

CodeVista analyzes your codebase and generates a stunning single-page HTML report — health scores, security scans, architecture detection, code smells, decay tracking, and more. All in pure Python.

The Problem

Every developer has stared at a codebase and thought: "I wish I could see the big picture." How healthy is this code? Where are the security holes? What's the architecture? Which files are decaying?

Existing tools are either:

Heavy (require Node.js, servers, or databases)
Expensive (SonarQube, CodeClimate)
Limited (only do one thing — lint, or coverage, or complexity)

I wanted something that does everything and generates a single HTML file you can share with anyone.

Enter CodeVista

CodeVista is an open-source code analysis and security scanner that supports 80+ languages with 38 lint rules. Install it, point it at your project, and get a beautiful interactive report.

pip install codevista
codevista analyze ./my-project/
# Open report.html — that's it. No server needed.

What It Does

🏥 Health Scoring (0-100)

Composite health score across 6 categories:

Readability
Complexity
Duplication
Coverage
Security
Dependencies

🔒 Security Scanner

Detects:

Hardcoded secrets (AWS keys, GitHub tokens, Stripe, API keys, passwords)
Dangerous functions (eval, exec, shell=True, pickle)
Private key detection
Severity scoring (critical/high/medium/low)

🏗️ Architecture Detection

Automatically identifies patterns from your project structure:

MVC, MVVM, Clean Architecture, Hexagonal
Microservices, Event-Driven, CQRS
Repository, Service Layer, Singleton, Strategy, Observer
Includes architecture quality scoring

👃 19 Code Smell Categories

Goes beyond typical linters:

God Classes, Long Parameter Lists, Feature Envy
Shotgun Surgery, Dead Code, Magic Numbers
Copy-Paste Code, Missing Error Handling
Inconsistent Naming, Boolean Parameters, and more

📉 Decay Detection

Tracks how your codebase degrades over time using git history:

Complexity growth trends
Coupling growth
Duplication growth
Debt velocity (technical debt per week)
12-week predictions via linear regression
Identifies inflection points where quality shifted

🧬 CodeDNA Fingerprinting

Generate a unique fingerprint of your codebase structure for tracking and comparison.

📊 25+ Subcommands

codevista analyze ./project/      # Full analysis
codevista quick ./project/        # Fast analysis (~3 seconds)
codevista compare ./v1/ ./v2/     # Compare two codebases
codevista watch ./project/        # Re-analyze on file changes
codevista smells ./project/       # Code smell detection
codevista architecture ./project/ # Architecture patterns
codevista code-age ./project/     # File age, churn, risk
codevista health ./project/       # Health score only
codevista security ./project/     # Security scan only
codevista dna ./project/          # CodeDNA fingerprint
codevista decay ./project/        # Architectural decay
codevista team ./project/         # Team productivity analysis
codevista trends ./project/       # Health trends over time
codevista lint ./project/         # Language-specific lint rules

👥 Team Analysis

Metric	Description
Lines per Author	Added/removed/net per developer
Bus Factor	People needed to understand 50% of code
Code Ownership	Contribution share
Onboarding Complexity	How hard for a new contributor to ramp up
Time Zone Distribution	When the team commits

Zero Dependencies. Seriously.

CodeVista is built with pure Python stdlib. No external packages. Install it from PyPI and it just works.

CI/CD Integration

Export to SARIF for GitHub Actions, Checkstyle, and more:

codevista ci-output ./project/ -f sarif
codevista export ./project/ --all

Why I Built This

As a student researcher working on multiple projects, I needed a way to quickly understand codebases — my own and others'. Commercial tools were too expensive, open-source tools were too narrow. So I built the tool I couldn't find.

Get Started

pip install codevista
codevista analyze ./your-project/

⭐ GitHub: github.com/rudra496/codevista
📦 PyPI: pypi.org/project/codevista

Rudra Sarker — student researcher and full-stack developer from Bangladesh. Building ethical, impactful open-source tools.

🔗 Connect:

GitHub | LinkedIn | Twitter/X | ResearchGate | Personal Site

🤟 I Built the Most Comprehensive Open-Source Sign Language Dataset Catalog — 73+ Datasets, 26 Languages

Rudra Sarker — Fri, 27 Mar 2026 10:52:07 +0000

How I curated and verified 73+ sign language datasets covering 26 sign languages to help researchers, developers, and the deaf community build better assistive technology.

Why I Built This

Sign language recognition (SLR) is one of the most impactful applications of computer vision and AI — it directly helps millions of deaf and hard-of-hearing people communicate. But when I started working on assistive technology projects, I hit the same wall every researcher faces: finding reliable, verified datasets is painfully hard.

Most existing lists are:

❌ Full of broken links
❌ Missing sample counts
❌ No license or citation info
❌ Limited to 1-5 sign languages

So I decided to build what I wish existed: a single, verified, comprehensive catalog.

What Is SignLanguage-Dataset-Hub?

SignLanguage-Dataset-Hub is a curated, open-source catalog of 73+ verified sign language datasets covering 26 sign languages — the most comprehensive collection available anywhere.

Key Numbers

Metric	Count
Verified Datasets	73+
Sign Languages Covered	26
Modality Types	Video, Image, Sensor, Pose, RGB-D, Skeleton, Text
URL Verification	100% (all links checked)
Tutorials Included	9 (beginner to advanced)

Languages Covered

The catalog spans sign languages from around the world:

ASL (American), BSL (British), ISL (Indian), CSL (Chinese), LSM (Mexican)
ArSL (Arabic), BdSL (Bangla), JSL (Japanese), KSL (Korean), TİD (Turkish)
DGS (German), LSF (French), NGT (Dutch), LIS (Italian), RSL (Russian)
Plus 11 more including multilingual corpora and linguistic databases

What Makes It Different

Every Link is Verified

Not just collected — verified. Every URL resolves to HTTP 200 or an auth-gated page. No placeholder links. No fabricated data. If we couldn't verify a dataset, it's excluded.

From Original Sources

Sample counts come from the original publications, not estimated or copied from secondary sources.

Full License & Citation Info

Every dataset includes:

Source URL (verified)
Sample count
License information
Proper citation to original creators

Working Tools Included

This isn't just a list — it comes with tools:

# Load demo sensor data (4,824 samples)
from scripts.data_loader import BdSLSensorGloveDataset
dataset = BdSLSensorGloveDataset(split='train')
print(f"Loaded {len(dataset)} samples, 36 gesture classes")

# Visualize
python tools/visualize.py --data data/bdsl/BdSL-Sensor-Glove/

# Query the catalog
import pandas as pd
df = pd.read_csv('datasets_catalog.csv')
asl = df[df['language_code'] == 'ASL']

9 Step-by-Step Tutorials

From beginner to advanced:

Introduction to Sign Language Recognition
Loading and Exploring Datasets
Visualizing Sign Language Data
Building Your First Classifier
Hand Pose Estimation with MediaPipe
Data Augmentation Techniques
Real-time Recognition System
Continuous Sign Language Recognition
Multilingual Sign Recognition

Project Structure

SignLanguage-Dataset-Hub/
├── DATASETS.md              # Complete verified catalog (73 datasets)
├── datasets_catalog.csv     # Machine-readable catalog
├── STATISTICS.md            # Detailed statistics & breakdowns
├── data/bdsl/               # Demo sensor dataset (4,824 samples)
├── docs/
│   ├── BENCHMARKS.md        # Published accuracy numbers
│   ├── TUTORIALS.md         # 9 tutorials (beginner to advanced)
│   ├── QUICKSTART.md        # Quick start guide
│   └── LICENSE_ATTRIBUTION.md
├── scripts/
│   ├── data_loader.py       # PyTorch data loaders
│   └── download_datasets.py # Multi-source downloader
└── tools/
    ├── visualize.py         # Sensor data visualization
    └── generate_realistic_data.py

Who Is This For?

Researchers — Find datasets for your next SLR paper with proper citations
Developers — Build sign language apps with verified, working data sources
Students — Learn sign language recognition with guided tutorials
The Deaf Community — Better tools start with better data

Why Open Source Matters Here

Sign language technology shouldn't be locked behind paywalls. The deaf community deserves free, accessible tools — and that starts with free, accessible data. This project is licensed under CC BY 4.0 so anyone can use it.

Get Started

git clone https://github.com/rudra496/SignLanguage-Dataset-Hub.git
cd SignLanguage-Dataset-Hub
pip install -r requirements.txt

🌐 Live Demo: rudra496.github.io/SignLanguage-Dataset-Hub
⭐ GitHub: github.com/rudra496/SignLanguage-Dataset-Hub

Citation

@misc{signlanguage_dataset_hub,
  title = {Sign Language Dataset Hub: A Verified Catalog of Sign Language Datasets},
  author = {Sarker, Rudra and Contributors},
  year = {2026},
  url = {https://github.com/rudra496/SignLanguage-Dataset-Hub}
}

I'm Rudra Sarker, a student researcher and full-stack developer from Bangladesh, building ethical AI and assistive technology. If you found this useful, ⭐ the repo and share it with someone who needs it!

🔗 Connect with me:

GitHub | LinkedIn | Twitter/X | ResearchGate | Personal Site

I Built a Tool to Validate AI-Generated Code (Now with JS/TS Support + AI Auto-Fix)

Rudra Sarker — Wed, 25 Mar 2026 11:38:34 +0000

We've all been there - you paste code from ChatGPT, Claude, or Copilot, it looks perfect, you ship it... and then it breaks because the AI invented a package that doesn't exist, or snuck in an eval(), or used an f-string for SQL queries.

84% of developers use AI coding tools now, but only 29% actually trust the output. That trust gap is real.

So I built AI Code Trust Validator - a tool that catches AI code problems before they reach production.

---

## What It Detects

**Security Vulnerabilities**
- SQL injection, command injection
- eval(), exec(), innerHTML XSS
- Hardcoded secrets, API keys
- Prototype pollution (JS)

**Hallucinations**
- Fake npm packages, invented Python modules
- Made-up functions and methods
- Placeholder API URLs

**Logic Errors**
- Infinite loops, unreachable code
- Missing await/async issues
- Bare except clauses, mutable defaults

---

## Supported Languages

- Python
- JavaScript  
- TypeScript

---

## Example Output

text
Analyzing: generated_code.py

TRUST SCORE: 67/100

Category Score Issues
Security 72 2 medium, 1 low
Hallucinations 45 3 critical
Logic 85 1 minor
Best Practices 70 2 warnings

Critical Issues:
[HALLUCINATION] Line 12: Import 'fancy_lib' does not exist
[SECURITY] Line 24: Potential SQL injection via f-string

Suggestions:
Replace 'fancy_lib' with 'numpy' or 'pandas'
Use parameterized queries


---

## Quick Start

bash
pip install ai-trust-validator
aitrust validate your_code.py --min-score 75


---

## What's New in v0.4.0

### AI-Powered Auto-Fix

Let GPT-4, Claude, or local Llama fix issues automatically:

bash
export OPENAI_API_KEY="sk-..."
aitrust ai-fix buggy_code.py --apply


Supports:
- OpenAI (GPT-4, GPT-3.5)
- Anthropic (Claude)
- Ollama (local models like Llama 3)

### JetBrains Plugin

Now works in IntelliJ, PyCharm, WebStorm with:
- Real-time code analysis
- Trust score in status bar
- One-click AI fixes

### JavaScript/TypeScript Support

Now validates .js, .jsx, .ts, .tsx files with:
- Security analysis (eval, innerHTML, prototype pollution)
- Hallucination detection (fake packages, invented functions)
- Best practices (var vs let/const, == vs ===)

---

## More Features

- VS Code extension
- Docker support
- GitHub Action for CI/CD
- Pre-commit hooks
- REST API server
- HTML/SARIF/JSON reports
- Team analytics dashboard

---

## Links

- GitHub: https://github.com/rudra496/ai-code-trust-validator
- PyPI: https://pypi.org/project/ai-trust-validator/
- Release Notes: https://github.com/rudra496/ai-code-trust-validator/releases/tag/v0.4.0

MIT licensed, open source. Contributions welcome!

---

What's the worst AI code hallucination you've encountered? Drop it in the comments and I'll add detection for it!