What Is Tolaria?
Tolaria is a desktop app for managing Markdown knowledge bases: a vault is a normal folder of Markdown files, and the app adds an editor, links between notes, saved views and a Git client on top. The repository describes it as a "desktop app to manage markdown knowledge bases" and the website as "a second brain for the AI era"; the author says he uses it for a vault of more than 10,000 notes.
The README names three uses: running a second brain or personal knowledge base, organizing company docs as context for AI, and storing the memory and procedures of AI assistants. The app is built with Tauri, React and TypeScript.
Get It
Tolaria is a free download for macOS, Windows and Linux, with a Homebrew cask on macOS; iPhone and iPad are in a TestFlight beta.
Option | How to get it |
|---|---|
| Desktop | Download page with all three desktop systems |
| Homebrew | brew install --cask tolaria (macOS) |
| iPhone and iPad | Beta via TestFlight; access is requested in the project Discord |
| Website | tolaria.md |
| Privacy policy | No standalone policy found; see the telemetry design note |
| Source code | GitHub: refactoringhq/tolaria, AGPL-3.0-or-later |
Version reviewed: stable 2026.10.13, released on 13 October 2026 (release page); alpha builds were being published the same day. The software is free and the website describes no paid tier.
Requirements
Tolaria publishes builds for the three major desktop systems; the sources state no RAM, disk or GPU requirement.
Operating systems
- What the sources say:
- macOS (Apple silicon and Intel), Windows x64 and Linux x86_64; macOS is the primary target and Windows and Linux are called "supported, early"
Installers
- What the sources say:
- A .dmg for macOS, a setup .exe for Windows, and AppImage, deb and RPM files for Linux
Hardware
- What the sources say:
- No minimum stated; any model inference runs in the agent or provider you connect, not in Tolaria itself
Git
- What the sources say:
- Optional: a plain folder opens as a vault, and Git adds history, commits, pull and push
Internet
- What the sources say:
- The README says a vault works completely offline; Git remotes and hosted model providers need a connection
The download page lists no minimum operating-system versions. The README says Windows installers are Authenticode-signed, while the install page says that signing is still to be added after certificate provisioning; company-managed devices may need IT approval first.
How to Get Started
The steps follow the project's install page; PromptQuorum has not run them.
- 1Install
Why it matters: Download the stable build for your system, or run the Homebrew cask on macOS. - 2Open a vault
Why it matters: Clone the Getting Started vault for a guided tour, or open an existing folder of Markdown files. - 3Use the command palette
Why it matters: Press Cmd+K on macOS or Ctrl+K on Windows and Linux to reach commands. - 4Connect AI (optional)
Why it matters: Add a model provider in Settings, or install a supported coding-agent CLI that Tolaria then detects.
Features Confirmed by the Sources
Every item below comes from the project's website, documentation or README; none has been independently tested.
- Markdown vault and relationships. Notes are Markdown files with YAML frontmatter; types are navigation lenses rather than enforced schemas, and notes link through wikilinks with autocomplete.
- Editor. A block-based editor with slash commands, a raw Markdown mode, whiteboards, media previews and table navigation; the documentation also has guides for spreadsheets and HTML blocks.
- Git client. Commit, push, pull, whole-vault and per-note history, diffs and conflict handling against any remote you connect.
- Custom views and multiple vaults. Saved views and sidebar sections, plus an option to load several vaults into one graph.
- AI agents and models. The AI panel streams CLI coding agents that the documentation lists as Claude Code, Codex, GitHub Copilot, OpenCode, Pi, Antigravity CLI, Kiro and Hermes Agent, and answers in chat mode from note context through Ollama, LM Studio, OpenAI, Anthropic, Gemini, OpenRouter or any OpenAI-compatible endpoint. Tolaria also exposes an MCP server that can search, read, create and update notes.
Data Flow and Privacy
The project describes Tolaria as files-first and offline-first: notes live in a folder you choose, there is no account, and the README says a vault works completely offline. Sync is Git against any remote you set up, with "zero dependency on Tolaria servers" according to the README.
AI is where data can leave the machine. Coding agents run on your computer under their own authentication, a hosted model provider receives the note context you send it, and Ollama or LM Studio keep that context local. Provider API keys are not stored in vault settings; you save them on the device or read them from an environment variable. Crash reporting is opt-in through a consent dialog on first launch (stack traces, app version, operating system and architecture, with no vault content or file paths, per the dialog text), and the source also has a separate analytics switch under Settings.
📌Note: These are the project's declarations and observations from its source code, not audit results; PromptQuorum did not inspect network traffic. No privacy-policy page was found on the website or the download page; the closest document is the repository's telemetry design note.
Who Should Use It
- Writers and researchers who want a long-lived Markdown archive that works with ordinary text tools and Git history.
- People using AI coding agents on their notes who want every change to show up as a reviewable Git diff.
- Developers and teams keeping docs as AI context who are comfortable with Git and want files they can version.
What We Could Not Verify
- Real-world use. PromptQuorum did not install Tolaria, so editor speed, stability and the author's statement about a vault of more than 10,000 notes are untested.
- AI quality and cost. The agents and models are third-party tools; their results, costs and what they send to their own providers were not examined.
- Platform polish. The project calls Windows and Linux support early, and the sources differ on Windows signing; macOS notarization is not mentioned.
- Mobile beta. The iPhone and iPad app and the TestFlight access process were not tried, and the documentation lists its feature set only briefly.
- Privacy beyond the code. No privacy policy is published, and the opt-in and path-scrubbing statements come from the repository, not an audit.
Competitors and Alternatives
SiYuan
- Platforms:
- macOS, Windows, Linux, iOS, Android
- Price / license:
- Free core, paid sync / AGPL-3.0
- Key difference:
- Block-level references and built-in AI agent collaboration; Tolaria centers plain Markdown files and Git
Articles about SiYuan (1)
- SiYuan Review: The Privacy-First, Block-Based Knowledge BaseUpdated September 18, 2026
Also mentioned in:
Blinko
- Platforms:
- Self-hosted web app (Docker)
- Price / license:
- Free / GPL-3.0
- Key difference:
- Self-hosted notes with RAG search through a configured LLM provider; Tolaria is a desktop app on local files
Articles about Blinko (1)
- Blinko Review: Self-Hosted AI Notes with RAG SearchUpdated September 19, 2026
Also mentioned in:
- Local Knowledge & Retrieval Tools Compared (2026): RAG, Document Chat, Vector Databases, Notes and SearchUpdated September 20, 2026
- Surf Review: Deta's Open-Source AI NotebookUpdated September 19, 2026
Both apps are in the PromptQuorum directory; confirm each app's current price, license and platforms on its own listing.
Frequently Asked Questions
Can I open a folder of Markdown files I already have?
Yes. The install guide says you can open an existing folder of Markdown files as a vault. If the folder is not a Git repository, Tolaria can initialize Git only when you ask, and it avoids broad folders such as Desktop, Documents or Downloads unless they are clearly dedicated vault folders.
What is the difference between Stable and Alpha?
Stable is promoted manually and is the channel the download page points to; Alpha follows every push to the main branch, so it gets features earlier but can be rougher. Several alpha builds were published within hours on 24 October 2026, and the documentation advises committing or pushing vault changes before switching channels.
Can an AI agent change my notes without asking?
Coding agents can edit files in the vault. The documentation describes a Vault Safe mode limited to file, search and edit tools and a Power User mode that can allow local shell commands scoped to the vault for agents that support it. Direct model chat gets no vault-write tools or shell access, and a Git-backed vault shows changes as diffs you can review and roll back.
Is there a mobile app?
An iPhone and iPad app is in beta through TestFlight, and access is requested in the project's Discord community. According to the documentation it offers a WYSIWYG editor, Git auto-commit and sync through iCloud or Files, while the AI panel and some power features are desktop-only so far.
What license is Tolaria under, and can I reuse the name?
The website says AGPL-3.0-or-later and the GitHub API reports AGPL-3.0. The site adds that the Tolaria name and logo are covered by a trademark policy, so check it before reusing the branding; this is not legal advice.
Who is behind Tolaria?
The website says it was created by Luca, the founder of Refactoring, after years of writing and note-taking, and the README says he built it for himself and shares it as open source. The README names Codacy, CodeScene, CircleCI and Unblocked as tools that support the project; no company or paid plan is described.
Verdict
Tolaria is an actively developed, free, open-source way to keep Markdown notes as plain files under Git and hand them to the AI agents or models you already use. Against that, Windows and Linux builds are described as early, the project is presented by one named maker, no privacy policy is published, and none of its speed or AI claims could be checked here. It suits people who already like plain files and Git and want AI help with reviewable diffs; anyone who needs a documented privacy policy, a mature mobile app or built-in AI should test it first or look elsewhere.
Sources
- Tolaria website — feature overview, AI paths, license and links, checked 24 October 2026.
- Tolaria download page — stable version 2026.10.13, platforms and Windows signing note, checked 24 October 2026.
- Tolaria on GitHub — README, documentation, release channels and telemetry design note; latest commit on 24 October 2026.
- Stable release v2026-10-13 — installer files and release date, checked 24 October 2026.