Configuration and Security

Configuration and Security

Configuration

Runtime behavior is controlled through extension settings and selected environment variables.

Important areas include:

• probe backend selection
• indexing behavior
• root and custom path setup
• metadata extraction tools
• AI model bootstrap and optional HuggingFace access
• remote access and write authorization

Security Model

The backend is designed with a compatibility-first local model and explicit controls for remote write access.

Key points:

• loopback writes remain allowed by default
• remote writes require explicit authorization
• strict auth mode is available through environment configuration

Privacy And Offline Clarification

Recent docs separate local AI inference concerns from remote write-security concerns.

Important distinction:

• the HuggingFace token is about model downloads and Hub access
• the Majoor API token is about protecting remote write operations to the local backend

This distinction matters because users often read "token" as "cloud AI upload". The project docs now explicitly separate those concerns.

Repository Security Policy

The repository also includes a standard GitHub security policy file, but the operational security model for the extension is described more concretely in the docs listed below.

Canonical Docs

• Privacy and Offline Guide
• Settings Configuration
• Security Environment Variables
• Threat Model
• Security Policy