Tolaria is a macOS application designed for managing large Markdown-based knowledge bases that can be used collaboratively by humans and AI agents. It treats a user’s notes as a structured, interconnected system of plain Markdown files that both people and AI can read, create, and edit.

The core idea behind Tolaria is to turn Markdown into a shared workspace where AI agents actively participate in knowledge management. Agents can create new notes, link them to existing ideas, and modify content over time, making the knowledge base continuously evolving and self-organizing.

Tolaria is built around the concept of an “LLM wiki,” where the entire workspace is transparent and interpretable by AI systems because it is based on simple, structured text files. This makes it ideal for agent-driven workflows where context, memory, and relationships between ideas matter.

The application is designed for power users who manage large volumes of structured information. It supports extensive note systems, linking, and collaboration between multiple AI agents operating on the same knowledge graph.

Tolaria is also positioned as an experiment in AI-assisted software development and knowledge work, with a strong focus on observability, transparency, and reproducibility of how AI agents interact with content.

Key features include:

• Markdown-based knowledge base management system
• AI agent collaboration on notes and documents
• Creation and linking of structured knowledge graphs
• Local macOS desktop application
• Designed for large-scale note systems and “LLM wiki” workflows
• Human + AI shared workspace for evolving knowledge
• Fully file-based (Markdown) for transparency and portability

Common use cases include:

• Building personal or team knowledge bases
• AI-assisted note-taking and knowledge management
• Creating structured “second brain” systems
• Collaborative AI-human documentation workflows
• Managing large interconnected note systems
• Experimenting with agent-driven knowledge graphs

Tolaria is developed by Luca Rossi (Refactoring HQ) as an open-source experiment in AI-native knowledge systems and collaborative agent workflows. It is also part of a broader exploration of AI-assisted software development practices.