🔍 Hister: Your Personal Search Engine

Hister is a privacy-centric search tool designed to index the specific web pages you visit and the local files you maintain. Instead of relying on external providers, it creates a full-text index of your own data, making it instantly retrievable via a web interface, a terminal, or an AI assistant using the Model Context Protocol (MCP).

🚀 Getting Started

To get Hister up and running, follow these steps:

Acquire the Binary: Download the version compatible with your OS from the latest releases.
Rename: Change the filename to hister (or hister.exe for Windows users).
Set Permissions (Linux/macOS only): chmod +x hister
Launch:
- Linux/macOS: ./hister listen
- Windows (PowerShell): .\hister.exe listen
Configure Browser: Navigate to http://127.0.0.1:4433 and install the extension for Chrome or Firefox.

Once the extension is active, simply browse the web. When you return to Hister and search for a phrase from those pages, your first results will appear.

Pro Tip: You can expand your index by importing browser history, adding local directories, or uploading specific files.

🛠 Alternative Installation Paths

Method	Command / Source
Homebrew	`brew install hister`
Docker	Use the provided `compose.yml` or `Dockerfile`
Nix	Available via `flake.nix`

🌟 Key Features

Hister is built for power users who value their data. ~~Forget about cloud-based tracking~~—everything stays under your control.

Privacy-First Architecture: Zero telemetry and no mandatory cloud dependencies.
Deep Indexing: It performs full-text indexing; it doesn't just look at URLs or page titles.
Seamless Integration: The browser extensions automatically save pages as you surf.
Advanced Query Language:
- Field filters & Phrases
- Wildcards & Negation
- Aliases & Result Prioritization
Semantic Search (Optional): Integrate your own embeddings endpoint to find documents based on meaning rather than just keywords.
Versatile Clients: Access your data via:
1. Web Browser
2. Terminal (TUI/CLI)
3. AI Assistants (via MCP)
Multi-Tenancy: Supports multiple users, ensuring that documents and search results remain isolated on shared servers.

🛡 Privacy & Logic

Hister operates on a simple premise: your data belongs to you.

Data Flow: The browser extension transmits page content only to your designated Hister server (except for favicons).
Storage: All indexes and documents are stored locally on your server.
External APIs: Semantic search is the only feature that sends text externally, and only to an endpoint you configure.

The Search Logic

Mathematically, Hister treats your personal knowledge base as a union of sets: $\text{Search Space} = \{ \text{Web Pages} \} \cup \{ \text{Local Files} \}$ The engine then applies a weighted retrieval function to return the most relevant results based on your query parameters.

System Architecture

💻 Development

If you wish to contribute or build from source, ensure you have Go 1.26, npm, and a C compiler (for CGO).

Build Process

# Clone the repository
git clone https://github.com/asciimoo/hister.git
cd hister

# Build the project
./manage.sh build

Development Mode

To enable hot-reloading for the web application and automatic Go backend rebuilds, use: npm run serve:app This utilizes a Vite development server and air for the backend.

🤝 Community & Legal

Connect: Find the community on Discord or IRCNet in the #hister channel.
License: This project is licensed under the AGPLv3 or any later version.
Resources:

Topics: #golang #privacy #personal-search #mcp-server #semantic-search #indexing