Hister provides private memory for AI assistants via self-hosted indexing and MCP

Hister is a self-hosted search engine that automatically indexes visited pages, local files, and documentation while storing offline previews to keep content searchable. It exposes an MCP endpoint, allowing local AI assistants to query this personal indexed material rather than relying on model memory or live web fetches.

• Automatically indexes visited web pages, local files, and documentation.
• Stores offline previews to maintain searchability without constant connectivity.
• Exposes an MCP endpoint for integration with local AI workflows.
• Functions as a private knowledge base for local LLMs.

The tool aims to serve as a useful private knowledge base for users running local models or MCP-based workflows.