What is Glia?
Glia is a privacy-focused ‘memory bridge’ designed to connect browser-based AI chat sessions with your local development environment. By capturing conversations from platforms like ChatGPT, Claude, and Gemini and storing them in a local SQLite database, it enables your IDE to query past technical decisions and context without relying on cloud-based sync.
Why Founders and Engineers Need It
Engineers often lose critical context when switching between the web interface of an LLM and their local code editor. Glia solves this by providing a unified, searchable, and privacy-compliant knowledge base that lives on your machine, ensuring that sensitive project data remains offline and under your control.
Key Features
- Local-First Storage: All data is stored locally in SQLite; zero reliance on cloud services.
- Seamless Integration: Includes an MCP server for native integration with Cursor and Claude Code.
- Intelligent Retrieval: Uses hybrid search (sqlite-vec and FTS5) combined with HyDE to improve the relevance of retrieved technical context.
- Privacy First: Aggressively redacts PII like API keys and tokens before saving data.
- Knowledge Graph Extraction: Automatically extracts structured facts from chats to provide richer context to the LLM.
How to Use It
Install the browser extension to automatically capture interactions. Once running, the local MCP server allows your IDE to fetch relevant snippets from your chat history based on the current file context, effectively creating a ‘second brain’ for your coding sessions.
Integrations
Cursor, Claude Code, Chrome, SQLite.
