146 entries
SecurityRepository:agent-substrate/substrate
Agent Substrate is an open-source Go runtime designed to run millions of agent workloads securely and efficiently on Kubernetes. It isolates untrusted code using microVMs or gVisor and snapshots idle agents to save resources.
5.4kstarsGo
From Video: GitHub Trending: Emerging Open-Source Tools for AI Agents and Developers
Repository:NVIDIA/OpenShell
NVIDIA OpenShell is an open-source runtime written in Rust designed to run autonomous AI agents in secure sandboxes. It enforces declarative policies at the kernel level and guards access to network routes and credentials.
16.6kstarsRust
From Video: GitHub Trending: Emerging Open-Source Tools for AI Agents and Developers
X post:Rohit Ghumare
In a viral post, Rohit Ghumare humorously highlights the knowledge gaps of 'vibe coders' regarding fundamental software engineering practices. He lists essential security vulnerabilities, cloud configurations, and testing methodologies that are frequently neglected when relying solely on AI prompts.
9.4kbookmarks549.2kviews
X post:Ian Livingstone
Ian Livingstone announced the first draft of the 'Agent Baseline Software Factory Reference Architecture'. Developed alongside community contributors, it offers an opinionated blueprint to secure increasingly autonomous software factories.
263bookmarks16.8kviews
X post:yishan
Open-source project PentAGI reached number one on GitHub's daily trending list. The tool uses multiple cooperating AI agents to launch penetration attacks against target systems without human intervention.
2.1kbookmarks103.6kviews
Link:Snyk
Snyk has introduced a six-module training curriculum titled AI Security Engineer Foundations. Delivered online and in person, the half-hour sessions address security across agentic workflows, MCP architectures, and AI-assisted development.
From Video: AI Security Engineer Foundations: Securing LLMs, Agents, and MCP Tools
Repository:snyk-workshops/vulnerable-student-assistant
A demonstration repository by Snyk Workshops featuring an intentionally vulnerable AI chat application built with FastAPI and the Gemini 2.5 Flash model. It is designed for testing and learning about prompt injection and sensitive data exposure.
4starsPython
From Video: AI Security Engineer Foundations: Securing LLMs, Agents, and MCP Tools
VideoVideo:CURT
In this workshop, Snyk Developer Advocate Javier Garza breaks down foundational AI security practices. Walking through the OWASP Top 10 for LLMs, he demonstrates direct and indirect prompt injection attacks, sensitive data disclosure, and supply chain risks, while detailing defensive architectures including system prompts, human approvals, database views, and MCP security scans.
Repository:morluto/rea
REA (Reverse Engineer Anything) is an open-source MCP server that enables AI agents to inspect runtime behaviour, reverse engineer applications, and analyse native binaries.
29.3kstarsTypeScript
From X post: REA: MCP Server for Agentic Reverse Engineering
Repository:microsoft/mxc
Microsoft eXecution Container (MXC) is a cross-platform sandboxed code execution framework designed for running untrusted workloads such as AI model output, plugins, and tools across Windows, Linux, and macOS. It wraps native OS containment mechanisms and VMs into a unified policy engine and typed SDKs.
1.8kstarsRust
From X post: Microsoft MXC: Open-Source Sandboxing for AI Agents and Untrusted Code
X post:Simon Willison
Simon Willison highlights Microsoft's newly open-sourced MXC (Microsoft eXecution Container). The cross-platform library enables sandboxed execution of untrusted code—such as model outputs, plugins, and tools—across Windows, Linux, and macOS.
545bookmarks45.5kviews
X post:Trending GitHub Repositories
Open-source tool REA connects AI agents to reverse engineering utilities via the Model Context Protocol (MCP), enabling local analysis of binaries, Electron apps, and web applications.
283bookmarks14.7kviews
Repository:bex-co/bex-security
Bex Security is an open-source, upstream-first fork of OpenAI Codex Security that brings vulnerability detection, validation, and remediation to multiple coding agents via the Agent Client Protocol (ACP).
51starsTypeScript
From X post: Cloudflare Releases Security Audit Skill for Coding Agents
Repository:martinholovsky/SOTA-skills
SOTA-skills is a collection of 42 modular engineering skills for AI coding agents such as Claude Code. It introduces build and audit rules via a dynamic loop rather than relying on bloated prompt dumps.
22starsPython
From X post: Alibaba Releases Open-Source Code Review Tool
Repository:nealbridges/VulnHunter
VulnHunter is an open-source, agentic AI security scanner that inspects source code from an attacker's perspective, confirms exploitability, and generates fixes. This maintained fork of Capital One's original project decouples the tool from Claude Code to support arbitrary agent harnesses and open-weight models.
223starsPython
From X post: VulnHunter: Open-source AI security scanner for vulnerability hunting
Repository:rauchg/gdp-ts
gdp-ts brings the 'Ghosts of Departed Proofs' paradigm to TypeScript, enforcing authorization and entitlement checks at compile time via the type checker.
550starsTypeScript
From X post: gdp-ts: Type-Level Authorization Proofs for Sensitive Operations and Agent Code
X post:Mia
Mia highlights VulnHunter, an open-source agentic security scanner that inspects source code like an attacker, validates flaws, and develops proofs of concept alongside suggested fixes.
229bookmarks18kviews
X post:Chris Tate
Chris Tate and Guillermo Rauch showcase gdp-ts, a TypeScript library implementing 'Ghosts of Departed Proofs' to enforce authorization checks at compile time via the type system.
99bookmarks16kviews
X post:Vaishnavi
Developer Vaishnavi has introduced REA (Reverse Engineer Anything), an open-source toolkit that brings reverse-engineering capabilities to AI agents via MCP and a CLI.
9kbookmarks292.5kviews
X post:Cantina
Cantina has introduced apex-flash-1, an open-weights model designed for security research. It was post-trained on real vulnerabilities identified by the team, alongside an abliterated version designed to reduce refusals in authorized workflows.
1.5kbookmarks270.4kviews
Repository:Lakr233/vphone-cli
vphone-cli creates and runs virtualized iOS instances on Apple Silicon Macs running macOS 15 or newer. The tool leverages Apple's Virtualization.framework and the PCC research VM infrastructure.
14.7kstarsSwift
From Video: Trending GitHub Projects: OpenBao, Magnitude, herdr, and Other Open-Source Tools
Repository:openbao/openbao
OpenBao is an open-source solution licensed under MPL-2.0 designed to manage, store, and distribute sensitive data such as credentials, certificates, and API keys. The project operates under open-governance principles supported by the OpenSSF.
8.2kstarsGo
From Video: Trending GitHub Projects: OpenBao, Magnitude, herdr, and Other Open-Source Tools
Link:stackone.com
StackOne has launched Falcon Engine, a runtime layer positioned between AI agent frameworks and enterprise software to handle tool execution, token optimisation, and security controls.
From X post: StackOne Falcon: Tool Calling Execution Engine for AI Agents
Repository:snyk/agent-scan
Agent Scan is an open-source security tool from Snyk that discovers and scans local AI agent components, MCP servers, and skills for prompt injections and vulnerabilities.
3.1kstarsPython
From Video: Benchmarking and Refining a Snyk AI Skill Against Anthropic Rules
24 of 146 shown
Do you have an interesting article, a video, or a tweet that belongs in our knowledge base? Write to us at office@webconsulting.at.
Parts of this content were created with the assistance of AI.
