Front FAQs

Question 1

What is the Front MCP Server?

Accepted Answer

The Front MCP Server is a secure, fully-featured Model Context Protocol (MCP) server that enables Large Language Model (LLM) agents to interact seamlessly with the Front Platform API. It offers 26 compound tools providing over 200 actions for reading, searching, managing, and acting on Front data.

Question 2

How does the Front MCP Server ensure security for LLM agents?

Accepted Answer

It implements a robust security model including HTTPS enforcement, AES-256-GCM token encryption with PBKDF2 key derivation, strict 0600 file permissions for tokens, configurable output and log redaction, and a policy engine that defaults destructive actions to 'deny'.

Question 3

What types of actions can LLM agents perform using this tool?

Accepted Answer

LLM agents can perform over 200 actions across 26 distinct categories such as managing conversations, contacts, messages, inboxes, tags, channels, and more. This allows for comprehensive interaction with Front data, from fetching information to creating, updating, and assigning resources.

Question 4

What authentication methods does Front MCP Server support?

Accepted Answer

The server supports both quick-start API token authentication and the recommended OAuth 2.0 method. OAuth provides enhanced security with automatic token refresh, encrypted local storage, and a user-friendly authentication flow via your browser.

Question 5

Can I control and limit the actions LLM agents can take?

Accepted Answer

Yes, a configurable policy engine allows you to define permissions. Actions are categorized into 'read', 'write', and 'destructive' tiers, each with a default decision (allow, confirm, deny). You can also set specific overrides for individual tools or actions, ensuring fine-grained control over LLM behavior.

Front

Front

主要功能

使用案例

主要功能

使用案例