WorkShadow

Like a shadow—records what you do, understands what you need.

WorkShadow is a local-first desktop work journal. Manage entries on the left and write in a rich-text editor on the right—daily work, decisions, issues, and notes. Data stays on your machine; for AI features (search, summaries, Q&A, image descriptions, and more), connect your own model services in Settings—you own the content, you choose the capabilities.

How is WorkShadow different from typical notes / docs tools?

Common pain point	What WorkShadow does
Easy to write, hard to find later	Keyword search + semantic retrieval—find details from months ago in one query
Digging through files for daily / weekly reports, emails, updates	Workspace · Log summary—select multiple logs and draft reports to your spec
“Why did we decide that back then?”	Workspace · Log Q&A—ask your history; get broader, connected answers
Worried about cloud lock-in or uploads	Data stays local; AI calls the network only when you configure models
Backup and migration	`.ws` full backup with merge import

WorkShadow is not a general-purpose Word or Notion replacement—it is a work journal companion for ongoing capture, review, and reporting: smooth writing, fast finding, easier summaries, with data and model choice always in your hands.

WorkShadow main window

Installer vs development build

The installer (Releases package) and development build (npm run tauri dev or dev installer) share the same core features and are both free. The installer is optimized for smoother, more fluid editing; it also adds smart completion and onboarding (✅ yes / ❌ no):

	Dev build	Installer	Notes
Price	🆓	🆓
Core features	✅	✅
Editing performance	❌	✅	Smoother, more fluid editing
More user-friendly, more polished UI & interactions	❌	✅	Smoother day-to-day experience
Image copy/paste & drag-and-drop	❌	✅	Paste or drag images into the editor
Smart completion	❌	✅	Local; smarter over time
Onboarding	❌	✅
Data on device	✅	✅
How to run	Download / from source	Download
Best for	Development	Daily use

Completion and onboarding learn on your machine—log bodies are not uploaded. AI features are enabled only when you configure models in Settings.

Features

Log organization & editing

Smart completion (installer only): Suggests continuations at the cursor based on logs already saved locally; learning and inference stay on-device—no body text upload. The more you use it, the better it matches your style. Not included in the development build.
Rich text: Headings, lists, task lists, quotes, code blocks, tables, links, images, video, math, and more.
Image copy/paste & drag-and-drop (installer only): Paste images from the clipboard or drag image files into the editor. Not included in the development build.
Batch actions: Multi-select nodes to move or delete in bulk.
Import Markdown: Bring existing .md files into a log entry and keep editing.

Search & understanding

Keyword search: Type in the left search box for local keyword matches with multiple snippets per log; click to open.
Semantic search: With an embedding model configured, search by intent via vectors—find similar meaning, different wording. Falls back to keyword search when not configured.

Workspace (AI-assisted; configure your own models)

Memory: Persistent notes across logs (OKR definitions, terminology, summary focus) used in summaries and Q&A.
Log summary: Select logs, combine Memory and writing preferences (focus, tone, structure) to draft daily / weekly / monthly reports, emails, status updates, project reports, and more—copy and tweak before sending.
Log Q&A: Retrieve relevant passages from all logs and synthesize an answer—more complete and coherent than reading one entry at a time; sources are cited.

Settings & data

UI: Light / dark theme; Chinese / English (or follow system); adjustable UI scale.
Paths: Custom log directory and temp directory; optional export of body text to your folder on save.
Models: Separate configs for LLM (summary, Q&A), multimodal (image / video description), embedding (semantic search); connection test supported.
Shortcuts: In-app shortcuts and system-wide “new child log” shortcut are customizable.
Import / export: Pack logs, memory, settings, and more into .ws backups; merge restore from backup.

Getting started (desktop)

Option 1: Installer

Visit the Releases page to download the installer (WorkShadow_*_x64-setup.exe). Launch from the Start menu or desktop shortcut after installation. Daily use:

1. First launch

Open WorkShadow.
Open Settings (top-right). Recommended:
- General: Theme, language, log save directory.
- Models (optional): For semantic search, summaries, Q&A, or image AI descriptions, enter base URL, API key, and model name, then test connection.
Back (top-left) to the main window.

2. Writing logs

Select a folder or log on the left; use New log below the search box.
Write in the editor; use the toolbar for tables, images, links, etc.
Save (or Ctrl+S / ⌘+S while focused in the editor): Writes to the local database, exports to your log directory per settings, and updates the search index.

3. Organize & find

Organize: Right-click to add children, rename, move, copy, delete; drag-and-drop; batch actions for multiple nodes.
Find: Search from the left box; with embedding configured, switch to semantic mode for natural-language retrieval.
Workspace: Open Workspace in the sidebar—Memory, Log summary, Log Q&A (summary and Q&A need an LLM configured).

4. Backup & migration

Under Settings → Data, export to a .ws file; on another machine, import from .ws to merge (categories not included in the export are left unchanged).

Option 2: Development build (installer or source)

You can also download the development build (WorkShadow_*_x64-dev-setup.exe) from Releases without setting up the dev environment below.

To run from source—for developers or anyone running the repo directly—prerequisites:

Node.js 18+ (current LTS recommended)
Rust + Cargo (Tauri)
On Windows: Visual Studio “Desktop development with C++” workload (Tauri prerequisites)

From the repo root:

npm install
npm run tauri dev

The first command installs frontend dependencies; the second starts the Vite dev server (default http://localhost:1420), then builds and opens the Tauri window loading that URL—frontend changes usually hot-reload without a full rebuild.

If port 1420 is in use, free it before retrying.

Note: npm run tauri dev is for local development and does not produce a distributable installer. For a release build, run npm run build then npm run tauri build; artifacts are usually under src-tauri/target/release/bundle/.

After it starts, daily use matches Option 1—follow “First launch” through “Backup & migration” above.

Contact us

Questions, feedback, or partnership inquiries? Reach us through any of the channels below. The same QR codes are also available in the app under Settings → About.

Administrator WeChat
_{Scan to connect for support and discussion}

WeChat official account
_{Scan to follow for product updates}

QQ group
_{Scan to join; group ID 1107536375}

Email: feiyangtech@qq.com

Open source & license

This repository is released under the GNU Affero General Public License v3.0 (AGPL-3.0).

You may use, study, modify, and distribute this software freely.
If you modify it and offer network access to it, you must provide corresponding source code to users (AGPL copyleft).
Third-party dependencies are subject to their respective licenses.