A developer has published Virgil, an open-source project that repurposes Claude Code as a daily journaling interface with memory that persists across sessions.
Claude Code, Anthropic's terminal-based coding assistant, doesn't retain memory between sessions by default. Each time you start a new session, it starts fresh with no knowledge of prior conversations. Virgil solves this by maintaining a structured journal of entries locally, giving Claude Code something to reference when you write a new entry or ask questions about past ones. The result is a journaling experience where the AI actually remembers what you wrote last week or three months ago - unlike most AI chatbots that forget everything the moment you close the window.
The project lives on GitHub under andrewspode/virgil. It's a thin layer on top of Claude Code rather than a standalone app, which means it inherits the full reasoning capabilities of the underlying model without requiring a separate AI subscription beyond what Claude Code already needs.
The use case is narrow but the approach is thoughtful. Most "AI with memory" products require routing your notes through someone else's cloud service. Virgil keeps everything local. For people who journal regularly and want an AI that can surface patterns in their thinking, reflect back what they wrote, or answer a question like "when did I last feel burned out?" - this is a more private alternative to cloud-based options.
It's also a good example of a pattern that's been picking up steam: people building personal productivity tools on top of coding assistants rather than consumer AI apps. Coding assistants tend to have better context management and direct file-system access, which makes them surprisingly capable outside of their intended use case.