Hister: a self-hosted, full-text search engine for your own browsing and files

Hister is an open-source personal search tool that indexes the full contents of the web pages you visit and the local files you keep, so you can retrieve things you’ve already seen without going back to a third-party search provider. It ships as a single binary that runs a local server, paired with Firefox and Chrome extensions that capture newly visited pages. Beyond automatic browser capture, it can import existing browser history, crawl websites, and index local directories, and it supports advanced queries with field filters, phrases, wildcards, negation, and result priorities.

The project’s central pitch is privacy through self-hosting: there is no telemetry and no mandatory cloud service, and indexed content stays on infrastructure you control. Data leaves your machine only in two optional cases—favicon downloads by the extension, and, if you enable semantic search, document text sent to an embeddings endpoint you configure. It also offers multi-user isolation for shared deployments.

Notably, Hister exposes its index through multiple front ends—a web UI, a terminal client, and an MCP server—letting an AI assistant query your personal corpus directly. Built in Go and licensed under AGPLv3, with installation via prebuilt binaries, Homebrew, Docker, or Nix, it positions itself as a locally controlled alternative to cloud search for personal knowledge retrieval.