
OpenShell
Sandboxed runtime for autonomous AI agents with declarative YAML policies enforcing filesystem, network, and process constraints, plus endpoint-bound…

Sandboxed runtime for autonomous AI agents with declarative YAML policies enforcing filesystem, network, and process constraints, plus endpoint-bound…

Post-exploitation framework for automated network authentication testing, credential harvesting, and lateral movement across Windows/AD environments…

PoC exploit for CVE-2026-41940 targeting cPanel/WHM authentication bypass via CRLF injection. Supports session hijacking, RCE, reverse shell, mass…

CVE-2026-6875 ServiceNow Pre-Auth RCE Framework 🔥 JS Injection → Sandbox Escape → RCE → Root. Features: --detect, --exec, reverse/interactive shell,…

Safety cannot be a prompt instruction. TBP provides an external execution-layer boundary for autonomous agents, enforcing hard F/I/W invariants via…

Rust CLI for MQTT penetration testing: subscribe/publish, broker fingerprinting, ACL probing, topic enumeration, credential brute-force, and Last…

Overlay network protocol giving AI agents permanent addresses, authenticated encrypted tunnels, and a trust model over UDP. Includes NAT traversal,…

Open-source security audits for Supabase: a read-only MCP server, a CLI agent and a Claude Skill. Finds RLS misconfigurations, exposed keys and risky…

Three unauthenticated vulnerabilities in the Thinkware U3000 dashcam's local WiFi control protocol: arbitrary file write, arbitrary file read, and…

Execution-Layer Security (ELS) for AI agents — policy-enforced shell with audit.

Home-lab penetration test report of Metasploitable3 covering Nmap recon, Drupalgeddon RCE, SQL injection, SSH credential reuse, sudo privilege…

Beacon Object File for in-line LSASS credential extraction using the KslD.sys BYOVD technique. Extracts NT hashes and cleartext passwords from…

A static + runtime security scanner for MCP (Model Context Protocol) servers

Self-contained Python PoC for Dovecot SQL authentication bypass: logs in as any user without the real password and enumerates usernames on vulnerable…

Lightweight CLI tool that runs AI coding agents inside isolated Bubblewrap sandboxes with strict filesystem, network, and credential isolation to…

CVE-2025-48932 - Unauthenticated SQL injection exploit for Invision Community ≤ 4.7.20. Fully automated exploitation with database enumeration,…

Critical authentication bypass exploit for cPanel/WHM CVE-2026-41940. Leverages CRLF injection in cpsrvd daemon to gain root WHM access without…

LLM-backed AI agent security — inbound injection detection + outbound privacy protection