Aguara FAQs

Question 1

What types of threats can Aguara detect?

Accepted Answer

Aguara features comprehensive threat coverage with 138 rules across 15 categories, including Prompt injection, Credential leak, Exfiltration, External download, Supply chain attacks, Command execution, and insecure MCP server configurations.

Question 2

Does Aguara require internet access or cloud services?

Accepted Answer

No, Aguara operates entirely locally and deterministically. It requires no network access, LLM calls, or cloud dependencies, ensuring fast, private, and consistent security scans within milliseconds.

Question 3

How is Aguara integrated into an AI agent's workflow?

Accepted Answer

Aguara integrates as an MCP server tool that AI agents can call. Agents use functions like `scan_content` for text analysis (e.g., skill descriptions) or `check_mcp_config` for server configurations, receiving structured reports on findings before making decisions.

Question 4

What kind of output does Aguara provide?

Accepted Answer

Aguara generates structured reports with severity-rated findings, matched patterns, and line numbers. This allows AI agents (or developers) to quickly understand detected threats and their context, enabling informed decisions to prevent security incidents.

Question 5

What is Aguara and how does it secure AI agents?

Accepted Answer

Aguara is an AI agent security advisor that scans skills, plugins, and Model Context Protocol (MCP) configurations for threats before installation or execution. It empowers agents to identify dangers like prompt injection, credential leaks, and supply chain attacks, preventing compromised operations.

Aguara

Aguara

Key Features

Use Cases

Key Features

Use Cases