goniszewski/grimoire

What it solves

Grimoire is a local-first bookmark manager designed for technical resources. It solves the problem of losing track of saved links by allowing users to save, organize, and search through their knowledge base privately on their own machine, avoiding reliance on cloud services.

How it works

The system consists of a React-based frontend and a Bun/Hono/SQLite daemon that runs locally. When a URL is saved, the daemon fetches the content, extracts readable text and metadata from sources like GitHub, YouTube, and PDFs, and stores it in a local SQLite database. Users can optionally connect AI providers (such as Ollama, OpenAI, or Anthropic) to automatically summarize content, generate tags, and create semantic embeddings for meaning-based search.

Who it’s for

It is primarily for people who save large amounts of technical documentation and resources and require a private, local-first tool for organizing and retrieving them.

Highlights

• Local-First Privacy: All data is stored locally in SQLite, ensuring user privacy and security.
• Hybrid Search: Combines SQLite FTS5 keyword search with semantic embedding search for more accurate retrieval.
• AI Enrichment: Optional integration with various AI providers for automated summarization and categorization.
• Broad Content Extraction: Specifically handles technical formats including GitHub repositories, issues, and YouTube transcripts.
• MCP Support: Exposes a Streamable HTTP Model Context Protocol (MCP) for local integrations.