Hister: A Private Search Engine for Personal Web Pages and Local Files

Introduction to Hister

Hister is a newly released private search engine focused on personal data. It indexes the complete content of web pages a user visits and files stored locally on their system. This allows users to search their own digital footprint for information they have previously encountered or stored.

Key Features and Functionality

The search engine provides full-text indexing, going beyond just titles and URLs to include the actual content of documents. It integrates with browsers like Firefox and Chrome via extensions to automatically save newly visited pages. Users can perform searches through a web interface, a terminal user interface (TUI), command line, or an AI assistant connected via MCP.

Hister supports powerful queries with field filters, wildcards, negation, and result priorities. An optional semantic search feature allows finding documents by meaning through a user-configured embeddings endpoint. It also includes multi-user support to keep data separate on a shared server.

Privacy and Installation

Hister emphasizes privacy, operating without telemetry or mandatory cloud services by default. Users can run it locally or on their own infrastructure. The browser extension sends indexed page content only to the user's configured Hister server. Optional semantic search requires configuring an embeddings endpoint, which may involve sending document text to a chosen service.

Installation methods include direct binary download, Homebrew, Docker, and Nix. Basic setup involves running the `hister listen` command and installing a browser extension to begin indexing and searching.