Aguara MCP is a local security scanner designed to protect AI agents (such as Claude Code, Cursor, Windsurf, or any MCP-compatible agent) from threats in untrusted third-party tools, skills, and configurations. It runs fully offline inside the agent's MCP server, providing rapid, structured security assessments for agent content, MCP settings, READMEs, and plugin manifests before any potentially dangerous action is taken. It detects issues like prompt injection, credential leaks, command execution, supply-chain risks, and more, giving agents a trust advisor before acting on external code.
Visit Aguara MCP's official website for product details and getting started.
Comprehensive user guide and API reference for Aguara MCP.
Community forum for reporting issues and discussing features related to Aguara MCP.
Examples and use cases demonstrating how to effectively utilize Aguara MCP.