CodeHub

21/10/2026

21/10/2026

🧠 OpenOSINT — AI-Powered Open-Source Intelligence

OpenOSINT is an open-source intelligence (OSINT) platform that helps security researchers investigate publicly available information using AI-assisted workflows. It combines multiple investigation tools with a command-line interface, interactive REPL, and MCP server for AI integrations.

✨ Features:

Provides 20+ tools for OSINT investigations

Supports email, username, and other public-information checks

Offers an interactive command-line interface

Integrates with AI assistants through an MCP server

Supports local models and a browser-based interface

🔗 GitHub Repository: in comments

21/10/2026

🛡️ Guardana — Open-Source AI Security Verification

Guardana is an open-source security tool that helps teams assess AI models, live AI endpoints, MCP servers, and recorded AI-agent activity. It provides reproducible security findings to help developers make better decisions before releasing AI-powered applications.

✨ Features:

Runs 58 built-in security checks

Scans code and AI model artifacts

Tests live AI endpoints and MCP servers

Analyzes recorded AI-agent ex*****on traces

Generates security evidence to support release decisions

🔗 GitHub Repository: in comments

21/10/2026

🔐 Pipelock — Open-Source AI Agent Firewall

Pipelock is an open-source security tool designed to protect AI agents from threats such as prompt injection, sensitive-data exfiltration, and server-side request forgery (SSRF). It monitors outbound traffic from AI agents and MCP integrations to help prevent malicious or unauthorized data transfers.

✨ Features:

Scans HTTP traffic for potential data leaks

Helps detect prompt-injection attacks

Protects against SSRF-related risks

Monitors MCP, A2A, and WebSocket traffic

Generates signed audit evidence for security investigations

🔗 GitHub Repository: in comments

20/10/2026

umm 😂 this can't be true

20/10/2026

haha 😂😂

19/10/2026

stop studying Cybersecurity the wrong way!

19/10/2026

Can a website really be compromised in just four minutes?

19/10/2026

at most times yeah 😂😂