You have probably already tried the alternatives.

You open your editor. The first thing you do is check whether the context file is current. It is not. The feature you shipped yesterday is not mentioned. The architecture decision you made on Tuesday is not reflected. So you spend 15 minutes updating a markdown file that your AI assistant will read once and then forget by the next session.

This is the workflow that every developer using AI coding tools has accepted as normal. Write the context file. Forget to update it. Watch the AI make wrong assumptions. Fix the assumptions. Update the file. Repeat.

Glia Guru reads your vault. It knows which tasks are open, which architecture notes you updated yesterday, and what you shipped last week. It generates a fresh CLAUDE.md, AGENTS.md, and tool-specific prompt files from your actual project state. When your project evolves, the context files update automatically. You never write CLAUDE.md by hand again.

The Obsidian plugin ecosystem is extraordinary. Tasks, Dataview, Projects, Kanban — each solves a specific problem within a single vault view. But none of them give you a cross-project dashboard. None generate AI context files. And none sync your project state to a database that a web dashboard or mobile app can read.

If you manage one project in one vault, plugins are often enough. The moment you manage multiple projects across machines, you are stitching together views from multiple plugins and hoping they agree.

Glia Guru is not a plugin. It is a standalone application that reads your vault externally, syncs the project state to your own Supabase account, and provides a visual dashboard that works on any device. Your vault structure stays the same. Your plugins keep working. Guru adds a layer on top — the AI context engine, the visual planner, and the cross-project dashboard sit outside Obsidian's plugin architecture by design.