
kavach-loop
Closed-loop vulnerability remediation: detect -> reproduce exploit -> generate patch -> verify with a proof harness. Worked example: Langflow…

Closed-loop vulnerability remediation: detect -> reproduce exploit -> generate patch -> verify with a proof harness. Worked example: Langflow…

CVE-2026-99999 Working PoC — real VM escape for Linux kernel csum_start bug (CVSS 9.3). Guest-to-Host heap leak. See also: CVE-2026-99999-analysis…

🔎 Static code analysis engine to find security issues in code.

Reachability triage for vulnerability scans. Feeds scanner findings to an LLM agent that reads your source and decides which CVEs are actually…

Executes untrusted Python, JavaScript, and Bash in a hardened jail with 5-second, 128 MB, 0.5-core, PID and output limits.

OPNsense 26.7.6 - High - Remote - Privilege leftover - custom firmware mirror sed-skips pkg fingerprints

Service that builds enrolled open-source repositories in isolated offline VMs and scans them for security vulnerabilities, emailing findings with…

Example AI-agent taskflows and MCP servers for auditing source code, triaging issues, and detecting vulnerabilities across web, Android, and iOS…

Multi-agent static application-security review harness for AI coding agents: maps codebases, hunts vulnerability classes, chains and verifies…

LLVM-based security research toolchain: NeverC, a C23 cross-compiler, and NeverD, a binary analysis and decompilation engine that lifts PE, ELF,…

Sandbox for AI coding agents. Runs Copilot CLI, Claude Code, OpenCode, Gemini CLI, Antigravity, Pi, goose or a plain shell inside a kernel-level…

Deliberately vulnerable Java/Maven fixture for testing Log4Shell (CVE-2021-44228) detection, code-impact classification, and remediation guidance in…

Security scanner auditing Claude Code environments for CVE-2026-21852 pre-trust execution, hook hijacking, and eBPF lockdown.

Proof-of-concept demonstrating a Node.js permission model bypass (CVE-2026-21636) that allows network access via undici/fetch to local services,…

Rootless container runtime and sandbox that launches kernel-enforced OCI images in milliseconds with no daemon, featuring resource profiles, seccomp…


Sandboxed devcontainer for running Claude Code in bypass mode safely. Built for security audits and untrusted code review.

Kernel-level eBPF sandbox for securing LLM agent tool calls made through the Model Context Protocol (MCP)