What Is Model Context Protocol?
Learn how MCP hosts, clients, servers, capabilities, transports, and downstream systems work, including the security responsibilities MCP leaves open.
Read articleAgentGuard Research
Security thinking for the agentic era.
Practical guidance, independent reviews, and clear explanations for teams building AI agents that can take real action.
What is Prompt Injection & How it Works
Prompt injection lets untrusted text override an AI system’s intended instructions. It may come directly from users or indirectly through webpages, documents, emails, and tools. Defenses should separate data from authority, restrict permissions, validate actions externally, require approval for sensitive operations, and log decisions. Simple phrase filtering is insufficient protection.
Read articleAgent Dependency Pollution: A Working Definition
An agent can make a bad security decision before it ever calls a tool: it can accept a package, plugin, skill, tool description, or MCP server that is not what its owner believes it to be.
Read article