Most project management tools ask you to move your work into their system. Glia Guru does the opposite. It comes to where your work already lives.
The tool started in late 2025 as a personal dashboard for managing a portfolio of web projects. Five codebases. Two machines. One Obsidian vault tracking everything from sprint items to architectural decisions. The vault was the source of truth, but it had no visual layer. No timeline. No board view. No way to see what mattered this week without grepping markdown files.
So I built one. A React dashboard that reads from the vault and writes back cleanly. Over four months, it absorbed features: a planner with day, week, and month views. An intelligence layer that surfaces knowledge gaps. A time tracking system powered by agent session data. A vault health monitor.
Then the AI agents arrived. Claude, Cursor, Windsurf. Each needed project context. Each had a different format. Maintaining CLAUDE.md files by hand became another maintenance tax on top of the plugin maintenance tax I had already escaped by building the dashboard.
The AI context engine was the answer: read the vault structure, understand the project state, generate the context files automatically. That was the moment the internal tool became a product.