State of MCP Server Security 2025: 5,200 Servers, Credential Risks, and an Open-Source Fix
Discover, secure and manage AI agents & NHIs
Maintain real-time inventory of all AI agents and NHIs, with context to understand risk and business usage.
Identify and remediate AI agents and NHIs with excessive privileges, vulnerable configurations, abnormal activity, and policy violations.
Provision secure-by-design AI agents with short-lived credentials, just-in-time, precisely scoped access, and policy at creation.
Set policy to resolve hygiene issues, reduce attack surfaces and prevent compliance violations.
Manage AI agents and their NHIs from provisioning to decommissioning.
Detect and respond to threats such as compromised credentials and out-of scope agent actions.
Control and enforce policies across your NHI attack surface.
Manage NHIs from provisioning to decommissioning.
Detect and respond to suspicious NHI activity & 3rd party breaches.
Centralized secret management across vaults & cloud.
Discover and assess third-party apps & vendors accessing your environment
The latest on AI agent & NHI threats, products stories and more
Meet Astrix at industry leading events
Watch on-demand sessions and expert insights
The latest company announcements and press
How our customers secure their NHIs with Astrix
Latest reports and whitepapers about NHI security
What is Agentic AI and related NHI risks
Core concepts, functional components, and technical capabilities
What are they & common vulnerabilities
How attackers exploit them & how to prevent it
The risks they pose & how to secure them
Definition, common use, and why they're important to secure
Tal Skverer (a.k.a. “reverser”) is the Security Researcher at Astrix Security, specializing in reverse-engineering complex threats against non-human identities (NHIs). Based in Tel Aviv and educated at the prestigious Weizmann Institute of Science, Tal delves into real-world cloud attacks analyzing malicious service principals, API token abuse, and emerging LLM-agent vulnerabilities .
An active contributor to the OWASP NHI Top 10 and frequent speaker at cybersecurity events like RSA and CSA, Tal turns his deep technical insights into accessible, hands-on analysis. His work empowers security teams to detect and mitigate threats stemming from machine identities—bridging the gap between academic rigor and practitioner impact.
State of MCP Server Security 2025: 5,200 Servers, Credential Risks, and an Open-Source Fix
Astrix Research Presents: Touchpoints Between AI and Non-Human Identities
Introducing the OWASP NHI Top 10: Standardizing Non-Human Identity Security
Sumo Logic: Compromised non-human identity leads to potential supply-chain exploits