Help Center

Your First Sort — Start to Finish

New to Librariness? This tutorial walks you through everything from installation to your first successfully sorted folder. Takes about 10 minutes.

Step 1 — Install and launch

Download the installer from librariness.com and run it. If Windows shows a SmartScreen warning, click More info → Run anyway — this is expected for new apps. Librariness will open automatically after installation.

Step 2 — Set up your AI provider

Librariness needs an AI to read and classify your files. Go to Settings → AI Provider and choose one:

• Free option: Choose OpenRouter and create a free account at openrouter.ai. Free models are available that cost you nothing per sort.
• Best quality: Use your own OpenAI, Anthropic, or Google API key for the most accurate results.

Paste your API key and click Save. Librariness will confirm the key is valid.

Step 3 — Add a source folder

Click Add Local Folder in the sidebar and pick a folder to organize. For your first sort, we recommend your Downloads folder — it's usually messy, easy to undo, and a great test case.

Step 4 — Pick a taxonomy

A taxonomy is the folder structure Librariness creates. Select Dewey Decimal (the default) to get started. It handles everything from business documents to photos to code files. You can customize it later.

Step 5 — Choose an output folder

Click Output Folder and select where you want the organized files to go. This can be a new empty folder or an existing one. Librariness will create subfolders inside it automatically.

Step 6 — Run it

Click Start Classification. You'll see a live progress view — each file being processed, what category it's being placed in, and the method used (keyword match vs AI). When it's done, browse your output folder to see the result.

The Taxonomy Tour v2.5

Librariness has a built-in guided tour that walks you through the taxonomy system — what each category means, how files get assigned, and what the output looks like. It's the fastest way to understand how the sorting logic works.

How to launch the tour

1. Open Librariness and go to the Taxonomy tab in the main panel.
2. Click the Taxonomy Tour button in the top toolbar.
3. Use the arrow buttons to step through each slide. You can exit at any time by closing the tour overlay.

What the tour covers

• Slide 1–2: Overview — what a taxonomy is and why the Dewey Decimal system works for files
• Slide 3: Top-level categories (000–900) and what kinds of files land in each
• Slide 4–5: Subcategories — how files get placed into specific subfolders within each category
• Slide 6: How the AI decides — keyword matching vs AI tier 1/2/3
• Slide 7: The Inbox — where uncertain files land and what to do with them
• Slide 8: Customizing — how to adjust categories to match your workflow

Building Custom Categories with the Dynamic Node Editor v2.5

The Dynamic Node Editor lets you visually build and edit taxonomy nodes — custom categories with their own rules, keywords, and AI context. You can use it alongside Danny the Librarian for real-time AI help while you build.

Opening the Dynamic Node Editor

1. Go to the Taxonomy tab.
2. Click the Dynamic Node Editor tab (next to the Standard tab in the taxonomy view).
3. Select an existing node from your taxonomy tree on the left, or click New Node to start fresh.

What you can do in the editor

• Name & code: Set the category name and Dewey code for the node
• Keywords: Add keywords that will trigger this category — files whose names or content match these words get routed here automatically without using AI credits
• AI context: Write a description of what belongs in this category. The AI reads this when classifying borderline files
• Subcategories: Add child nodes directly from within the editor

Using Danny while you edit

Danny the Librarian is available directly inside the Dynamic Node Editor — you don't have to switch panels. The Danny chat panel appears on the right side of the editor. You can ask him things like:

• "What keywords should I add to catch invoice files?"
• "Write me an AI context description for a Marketing Materials category"
• "What subcategories should this node have?"

Danny understands the context of what node you're editing and tailors his suggestions accordingly.

Resizing the Danny panel

Drag the bar at the top of the Danny panel up or down to resize it between about 140px and 650px. This lets you give more screen space to the editor or the chat as needed. The collapse arrow on the left of Danny's header bar hides/shows the panel entirely.

Setting Up Routing Rules v2.5

Routing rules let you override the AI's decision for specific files or patterns. If you always want files with "Invoice" in the name to go to a specific folder regardless of what the AI thinks, a routing rule handles that instantly without spending any API credits.

How to create a routing rule

1. Go to Settings → Routing Rules.
2. Click Add Rule.
3. Set the condition — this is what you want to match. Options include: filename contains, filename starts with, filename ends with, file extension is, or file path contains.
4. Set the destination — pick the taxonomy category (and subcategory) where matching files should always go.
5. Click Save Rule. The rule is active immediately for the next sort.

Rule examples

• All .pdf files → Documents / PDFs — useful if you want every PDF in one place regardless of content
• Filename contains "Invoice" → Finance / Invoices — catches invoices before the AI even looks at them
• File path contains "Screenshots" → Technology / Screenshots — routes an entire source subfolder to a fixed destination
• Filename ends with "_final" → Active Projects / Finals — useful for design or video workflows

Rule priority

Rules are checked in order from top to bottom. The first matching rule wins. Drag rules up or down in the list to change their priority. A file that matches multiple rules only follows the first one.

Getting the Most from Danny the Librarian

Danny is your AI assistant built into Librariness. He's not just a search tool — he can help you design your taxonomy, find files, explain classification decisions, and suggest routing rules. Here's how to get the most out of him.

Finding files

Danny searches across all your connected sources by filename and content. Some examples that work well:

• "Find my contract with Acme" — searches by name and content
• "Show me everything I have about machine learning" — content-based search
• "Any invoices from last year?" — combines date and category context
• "Where did my tax returns end up?" — finds by category and likely location

Understanding sort results

After a sort, you can ask Danny to explain specific decisions:

• "Why did 'project_brief.docx' go to 600 Technology?"
• "Which files were flagged for review?"
• "What went into the Inbox and why?"

Designing your taxonomy

Danny is great at taxonomy design, especially inside the Dynamic Node Editor. Ask him:

• "What subcategories make sense under Marketing Materials?"
• "Write an AI context description for a Legal Documents node"
• "Suggest keywords for catching HR-related files"

Where Danny appears

• Standard Taxonomy view — Danny panel on the right, collapsible
• Dynamic Node Editor — Danny panel built in alongside the editor (new in v2.5)

In both views, drag the top bar of the Danny panel to resize it. The collapse arrow is on the left side of the Danny header.

Sorting for Free with OpenRouter

You don't need to pay for AI to use Librariness. OpenRouter provides access to free AI models — including capable open-source models — that you can use for classification at no cost.

Setting up OpenRouter

1. Go to openrouter.ai and create a free account. You'll need to verify your email address.
2. In OpenRouter, go to Keys and create a new API key. Copy it.
3. In Librariness, go to Settings → AI Provider and select OpenRouter from the dropdown.
4. Paste your OpenRouter key and click Save.
5. In the model selector, choose a free model (models marked with a $0 cost per token are free). Good free options include mistralai/mistral-7b-instruct and meta-llama/llama-3-8b-instruct.

System Requirements

Librariness is currently available for Windows 10 and Windows 11 (64-bit). macOS support is planned for a future release.

Minimum Requirements

• OS: Windows 10 (64-bit) or newer
• RAM: 4 GB (8 GB recommended for large libraries)
• Disk Space: ~250 MB for the application
• Internet: Required for cloud sources and AI classification (unless using a fully local AI setup)
• AI Key: Required for classification. Free options available via OpenRouter — no credit card needed.

Installing Librariness

1. Download the installer from librariness.com — the file is named Librariness-Setup.exe.
2. Double-click the installer. If Windows shows a SmartScreen warning saying "Windows protected your PC," click More info → Run anyway. This is expected for new applications and is safe to proceed through.
3. The installer completes automatically in about 30 seconds. Librariness launches when done.
4. On first launch, the backend service starts up — this takes 5–10 seconds. You'll see the main interface once it's ready.

Running Your First Sort

Your first sort takes about 2 minutes to set up. See the full walkthrough in the First Sort tutorial above, or follow these quick steps:

1. Add a source: Click "Add Local Folder" in the sidebar. Start with your Downloads folder.
2. Set up an AI provider: Settings → AI Provider. Paste your key. OpenRouter has free models — see the Free AI tutorial.
3. Choose a taxonomy: Select Dewey Decimal or define a custom one.
4. Choose an output folder: Click Output Folder to set where organized files go.
5. Start Classification: Click the button and watch the live progress feed.

How long does a sort take?

• Under 500 files: 2–5 minutes
• 500–2,000 files: 5–20 minutes
• 2,000+ files: 30 minutes or more — let it run in the background

Keyword-matched files are instant. AI-classified files take a few seconds each depending on your provider. Routing rules run before everything else and are essentially instant.

Connecting Google Drive

1. In Librariness, go to Settings → Connected Accounts → Add Account → Google Drive.
2. A browser window opens asking you to sign in to Google. Sign in with the account that has the Drive you want to connect.
3. Google shows a permissions screen — click Allow.
4. You'll be redirected back to Librariness. Your Drive appears in the sidebar.

Troubleshooting Google Drive

• "This app isn't verified": Click Advanced → Go to Librariness (unsafe). This appears for apps in early access — it is safe.
• Browser doesn't redirect back: Close the tab and check if the connection already shows in Settings.
• Wrong account connected: Disconnect in Settings and reconnect, choosing the correct account when the browser opens.

Connecting OneDrive

1. Go to Settings → Connected Accounts → Add Account → OneDrive.
2. A Microsoft sign-in page opens. Sign in with your Microsoft account.
3. Review the permissions and click Accept.
4. Your OneDrive appears in the Librariness sidebar.

Connecting Box

1. Go to Settings → Connected Accounts → Add Account → Box.
2. Sign in to Box in the browser window that opens.
3. Click Grant Access to Box.
4. Your Box account appears in the Librariness sidebar.

Taxonomies Explained

A taxonomy is the folder structure Librariness uses to organize your files. Think of it as the blueprint — you define the categories, and Librariness fills them.

The Default (Dewey Decimal)

The built-in taxonomy is inspired by the Dewey Decimal Classification system. It organizes files into broad categories:

• 000 — General Works (software, databases, reference)
• 100 — Philosophy & Psychology (personal journals, self-development)
• 300 — Social Sciences (business, law, government)
• 500 — Science (research, data, technical papers)
• 600 — Technology (manuals, engineering, medical, IT)
• 700 — Arts (design files, photos, creative work)
• 800 — Literature (writing, scripts, documents)
• 900 — History & Geography (archives, records)

Each category contains subcategories for more precise placement.

Custom Taxonomies

If the default doesn't fit your workflow, build your own. Go to Settings → Taxonomies → New Taxonomy. Common approaches:

• By client: Client A / Client B → Invoices, Contracts, Deliverables
• By year: 2024 / 2025 / 2026 → by project type
• By project: Active / Archived / Templates

Use the Dynamic Node Editor to build and fine-tune custom nodes with keywords and AI context.

Dynamic Node Editor v2.5

The Dynamic Node Editor is a visual workspace for building and refining taxonomy nodes. Unlike the standard taxonomy tree view, the dynamic editor gives you a full editing canvas for each node — with keyword management, AI context fields, and Danny the Librarian built right in.

Opening the editor

Go to the Taxonomy tab and click Dynamic Node Editor in the tab strip. Select any node from the tree on the left to start editing it, or click New Node to create one from scratch.

Node fields

• Name & code: The display name and classification code (e.g. "650 — Business")
• Keywords: Exact words or phrases. Files whose names or content contain these are instantly routed here — no AI call needed
• AI context description: A plain-English description of what belongs in this node. The AI reads this for borderline files. The more specific, the better the accuracy.
• Subcategories: Child nodes nested under this one

Danny in the Dynamic Editor

Danny's chat panel appears on the right side of the Dynamic Node Editor. Ask Danny to help write keyword lists, AI context descriptions, or suggest subcategory structures. The editor and Danny share context — Danny knows which node you're working on.

Resize the Danny panel by dragging the bar above it. Collapse it entirely using the arrow on the left of the Danny header.

Routing Rules v2.5

Routing rules let you define exact filing instructions that override the AI. Any file matching a rule goes straight to the destination you specify — instantly, without using API credits.

Creating rules

Go to Settings → Routing Rules → Add Rule. Set a condition (filename contains, file extension, path contains, etc.) and a destination category. Rules are checked in order — the first match wins.

When to use routing rules

• You have a naming convention and want to enforce consistent filing (e.g. all files with "INVOICE" in the name)
• You have a folder of a specific file type that always belongs in one place
• You want to override AI decisions for specific patterns you know better than the AI does
• Speed — routing rules are near-instant and save AI credits for genuinely ambiguous files

See the full Routing Rules tutorial for step-by-step setup and examples.

Danny the Librarian

Danny is your AI-powered library assistant built into Librariness. He can find files, explain sort decisions, help design your taxonomy, and suggest improvements to your setup.

Where Danny lives

• Standard view: Click the chat icon in the sidebar, or find the Danny panel on the right side of the taxonomy view
• Dynamic Node Editor: Danny appears built into the editor panel (new in v2.5)

In both views, drag the bar above Danny's panel to resize it. The collapse arrow on the left of the Danny header bar hides/shows the panel.

What to ask Danny

• "Find my contract with Acme Corp"
• "Show me all invoices from last year"
• "Why did this file go to 600 Technology?"
• "What keywords should I add to catch HR files?"
• "Which files were flagged as needing review?"

Scanning Images & Scanned PDFs (OCR) v2.5

Librariness can read the text inside images and scanned PDF documents using OCR (optical character recognition). This means a photo of a receipt, a scanned contract, or a screenshot of a document can be classified by its actual content — not just its filename.

Supported file types

• Images: .png, .jpg, .jpeg, .tiff, .bmp, .gif
• Scanned PDFs: .pdf files that are images rather than selectable text

How it works

OCR is bundled directly into Librariness — it runs completely on your computer, no internet connection required, and your file contents never leave your machine. When Librariness encounters an image or scanned PDF, it automatically runs OCR to extract text, then classifies based on that text exactly like a regular document.

OCR accuracy

OCR works best on clean, well-lit scans with standard fonts. Handwritten notes, low-resolution photos, and skewed scans may produce less accurate text. If a file's OCR text isn't useful, Librariness falls back to classifying it by filename and file type.

Setting Up Your AI Key

Librariness uses AI to read and classify your files. You provide your own API key — your files go directly from your machine to the AI provider. Librariness never sees or stores your file contents.

Supported Providers

• OpenRouter — access to dozens of models including free ones. Best starting point. Create a free account at openrouter.ai. Verify your email before using or you'll get a 403 error.
• OpenAI (GPT-4o) — fast, widely tested. Get a key at platform.openai.com.
• Anthropic (Claude) — excellent for complex document understanding. Get a key at console.anthropic.com.
• Google (Gemini) — good alternative with a free tier. Get a key at aistudio.google.com.

Adding your key

1. Go to Settings → AI Provider.
2. Select your provider from the dropdown.
3. Paste your API key into the field.
4. Click Save. Librariness will test the key and confirm it's working.

How much does it cost?

Sorting 1,000 files typically costs between $0.05–$0.50 in API credits depending on the provider and model. Routing rules and keyword matches don't cost anything. Danny searches are very low cost. Use the OpenRouter free tier to pay nothing at all.

Free Trial

Librariness includes a 14-day free trial with full access to all features. No credit card required to start.

How to start your trial

When you first open Librariness, you'll be offered a free trial. Click "Start Free Trial" — the trial starts immediately and runs for 14 days.

What happens when the trial ends

Sorting and Danny will pause. Your existing library stays exactly as-is — nothing is deleted or moved. Subscribe to continue using the app.

Can I restart the trial?

No — the trial is once per machine. If you need more evaluation time, email admin@librariness.com and we'll work something out.

Activating a License Key

1. Purchase a subscription from librariness.com. You'll receive a license key by email from LemonSqueezy.
2. Open Librariness and go to Settings → License.
3. Paste your license key and click Activate.
4. Librariness verifies the key. Your subscription status updates immediately.

Moving to a new computer

1. On your old machine: Settings → License → Deactivate.
2. Install Librariness on the new machine and activate with the same key.

Billing & Cancellation

Subscriptions are managed through LemonSqueezy. View, change, or cancel at any time through your customer portal.

How to cancel

1. Find your original purchase email from LemonSqueezy.
2. Click the customer portal link in that email.
3. Click Cancel Subscription. Access continues until the end of the current billing period.

Refund policy

We don't offer refunds for partial billing periods. If Librariness isn't working for you, please email us before cancelling — we'd love the chance to fix it.

Windows Security Warning on Install

When you run the installer, Windows may show: "Windows protected your PC — Microsoft Defender SmartScreen prevented an unrecognized app from starting."

Why this happens

This appears for any app from a publisher without a code signing certificate. It does NOT mean the app is malicious — Windows shows this for new apps that haven't yet built up reputation data.

How to proceed

1. Click More info on the SmartScreen dialog.
2. Click Run anyway.
3. The installer proceeds normally.

Sort Errors

"API key invalid" or "AI provider error"

• Go to Settings → AI Provider and verify your key — no extra spaces.
• Check that your account has credits and isn't rate-limited (log in to your provider's dashboard).
• For OpenRouter specifically: make sure you've verified your email address — unverified accounts get a 403 error on every API call.

Sort stops mid-way

• Check your internet connection — required for AI classification.
• For cloud sources, check that your account is still connected (Settings → Connected Accounts).
• Very large files (>50 MB) can occasionally time out. Try excluding them from the sort for now.

Files not moving after sort completes

• Make sure you're not in Preview mode — the Sort button (not Preview) actually moves files.
• Check that Librariness has write permission to the output folder.
• On cloud sources, verify the cloud integration still has its permissions in the provider's app settings.

"Backend never responded" on startup

This was a known issue fixed in v2.5.1. Make sure you're on v2.5.2 or later (check Settings → About). If you're on the latest version and still seeing this, restart your computer and try again. If it persists, email support.

Cloud Authentication Issues

Google Drive says "This app isn't verified"

Expected during early access. Click Advanced → Go to Librariness (unsafe). The app is safe — Google shows this for apps that haven't completed full verification yet.

OneDrive "Need admin approval"

Your Microsoft account is managed by an organization with restricted third-party app access. Ask your IT admin to approve Librariness, or use a personal Microsoft account instead.

Auth window closes but nothing happens

1. Close and reopen Librariness.
2. Go to Settings → Connected Accounts — the account may already be listed.
3. If not, try connecting again. If the issue persists, email us with a description.

Still Need Help?

We're a small team that genuinely cares about making Librariness work for you. If you're stuck, reach out directly:

• Email: admin@librariness.com
• Response time: Typically within 24 hours on business days

It helps to include:

• Your Windows version (Win 10 or Win 11)
• Which version of Librariness you're on (Settings → About)
• What you were doing and what happened instead
• Any error messages or screenshots