Find a file
2025-11-24 00:40:29 +09:00
backups Refactor AI generation code and add DB utilities 2025-11-07 21:54:38 +09:00
node_modules Refactor AI generation code and add DB utilities 2025-11-07 21:54:38 +09:00
src fix: enhance user activity tracking by adding more event listeners to prevent interruptions during editing 2025-11-24 00:40:29 +09:00
tests Implemented backend support for a compendium feature, including database schema updates and frontend integration. 2025-11-08 01:45:12 +09:00
.eslintrc.json Refactor AI generation code and add DB utilities 2025-11-07 21:54:38 +09:00
.gitattributes Initial commit (code only; models ignored) 2025-11-07 04:24:59 +09:00
.gitignore fix: Update .gitignore to include todo.md 2025-11-17 18:43:44 +09:00
BACKUP_FEATURE.md feat: Implement GitHub Gists auto-backup feature with settings and restore functionality 2025-11-17 01:41:18 +09:00
BACKUP_QUICKREF.md feat: Implement GitHub Gists auto-backup feature with settings and restore functionality 2025-11-17 01:41:18 +09:00
BACKUP_TESTING.md feat: Implement GitHub Gists auto-backup feature with settings and restore functionality 2025-11-17 01:41:18 +09:00
favicon.ico Saving, export and compendium features updated, some UI changes added 2025-11-13 09:27:32 +09:00
ggml-base.dll Initial commit (code only; models ignored) 2025-11-07 04:24:59 +09:00
ggml-cpu-alderlake.dll Initial commit (code only; models ignored) 2025-11-07 04:24:59 +09:00
ggml-cpu-haswell.dll Initial commit (code only; models ignored) 2025-11-07 04:24:59 +09:00
ggml-cpu-icelake.dll Initial commit (code only; models ignored) 2025-11-07 04:24:59 +09:00
ggml-cpu-sandybridge.dll Initial commit (code only; models ignored) 2025-11-07 04:24:59 +09:00
ggml-cpu-sapphirerapids.dll Initial commit (code only; models ignored) 2025-11-07 04:24:59 +09:00
ggml-cpu-skylakex.dll Initial commit (code only; models ignored) 2025-11-07 04:24:59 +09:00
ggml-cpu-sse42.dll Initial commit (code only; models ignored) 2025-11-07 04:24:59 +09:00
ggml-cpu-x64.dll Initial commit (code only; models ignored) 2025-11-07 04:24:59 +09:00
ggml-rpc.dll Initial commit (code only; models ignored) 2025-11-07 04:24:59 +09:00
ggml.dll Initial commit (code only; models ignored) 2025-11-07 04:24:59 +09:00
libcurl-x64.dll Initial commit (code only; models ignored) 2025-11-07 04:24:59 +09:00
libomp140.x86_64.dll Initial commit (code only; models ignored) 2025-11-07 04:24:59 +09:00
llama-server.exe Initial commit (code only; models ignored) 2025-11-07 04:24:59 +09:00
llama.dll Initial commit (code only; models ignored) 2025-11-07 04:24:59 +09:00
logo.png Updated the README to reflect the main features 2025-11-16 03:32:54 +09:00
main.html feat: implement multi-tab synchronization and updatedAt timestamps for project, chapter, scene, and content management 2025-11-23 22:02:48 +09:00
mtmd.dll Initial commit (code only; models ignored) 2025-11-07 04:24:59 +09:00
package-lock.json Refactor AI generation code and add DB utilities 2025-11-07 21:54:38 +09:00
package.json Refactor codebase and add UI tests 2025-11-08 00:35:26 +09:00
README.md fix: Update model download links in installation instructions 2025-11-16 18:01:10 +09:00
start.bat feat: Enhance local model handling and user notifications in UI and scripts 2025-11-16 06:27:34 +09:00
start.sh fix: Update build timestamp in update checker module 2025-11-16 15:25:24 +09:00
W1-IMPORT-TROUBLESHOOTING.md Add W1 Import Troubleshooting Guide with common issues and solutions 2025-11-16 21:41:45 +09:00

Writingway Logo

Writingway 2

AI-powered creative writing tool — simple, friendly, and built for storytellers


🌟 What is Writingway 2?

Writingway 2 is a simple, powerful, and beginnerfriendly creative writing application designed for writers, not programmers.

It runs entirely in your browser, requires no installation, and stores all your work locally on your computer.
Just download → unzip → doubleclick start.bat → begin writing.

Writingway 2 is the successor to the original Writingway (Python/QML version).
Unlike its predecessor, this new version is built with JavaScript + HTML, making it lightweight, portable, and extremely accessible to users with minimal technical experience.


Key Features (Quick Overview)

  • 🚀 AIassisted writing tools

    • continue drafts
    • rewrite passages
    • improve style or pacing
    • brainstorm ideas
  • 📚 Compendium for worldbuilding

    • characters, places, lore, items
    • searchable entries
    • AIassisted creation
  • 🧩 Scenebased writing

    • organize by scenes & chapters
    • draganddrop reordering
    • autosave
  • 🧠 Workshop Chat

    Design prompts for and chat with AI about

    • plot generators
    • character builders
    • setting ideas
    • prose improvement
  • 🧰 Smart Prompt Builder (Scene Beats)

    • builds structured prompts automatically
    • integrates compendium + scene metadata
    • ensures consistent AI responses
  • 📥 Writingway 1 Importer

    • converts old Writingway projects
    • imports compendium & scenes
  • 🔌 Flexible AI backend

    • works with local LLMs (LM Studio, Ollama, Jan)
    • supports OpenAI / OpenRouter
    • automatic retry handling

📦 Installation

Windows:

  1. Download the latest ZIP release
  2. Extract the folder anywhere
  3. Download llama.cpp (for local AI):
    • CUDA (NVIDIA GPU) - Download llama-*-bin-win-cuda-cu12.2.0-x64.zip
    • CPU only - Download llama-*-bin-win-avx2-x64.zip
    • Extract llama-server.exe to the Writingway folder
  4. Download a model (optional, for local AI):
  5. Double-click start.bat
  6. Writingway opens at http://localhost:8000

Mac:

  1. Download the latest ZIP release
  2. Extract the folder anywhere
  3. Download llama.cpp (for local AI):
  4. Download a model (optional, same as Windows step 4)
  5. Run in Terminal:
    chmod +x start.sh
    ./start.sh
    
  6. Writingway opens at http://localhost:8000

Linux:

  1. Download the latest ZIP release
  2. Extract the folder anywhere
  3. Download llama.cpp (for local AI):
    • Ubuntu/Debian x64 - Download llama-*-bin-ubuntu-x64.zip
    • Extract llama-server to the Writingway folder
  4. Download a model (optional, same as Windows step 4)
  5. Run in Terminal:
    chmod +x start.sh
    ./start.sh
    
  6. Writingway opens at http://localhost:8000

Alternative: Use Cloud AI (No Local Model Needed)

If you don't want to download models:

  1. Skip steps 3-4 above
  2. Run start.bat (Windows) or ./start.sh (Mac/Linux)
  3. When prompted, choose to start without a local model
  4. In Writingway, go to ⚙️ AI Settings and configure:
    • OpenRouter (has free models like Gemini 2.0 Flash)
    • Claude API (Anthropic)
    • OpenAI API (ChatGPT)
    • Google AI (Gemini)

Advanced Users

You can also use:

  • LM Studio - Run models with a GUI, connect via API mode
  • Ollama - Multiple models, simple CLI interface
  • Jan - Cross-platform local AI app

Just configure the API endpoint in Writingway's AI Settings.


📝 Getting Started

1. Create a Project

Choose a project name. Writingway handles the rest.

2. Add Scenes

Scenes can be long or short.
Reorder them visually via draganddrop.

3. Use AI to Write or Improve

The AI panel allows you to:

  • continue your writing
  • polish prose
  • add detail
  • rewrite in a different tone
  • generate new ideas

4. Expand Your Worldbuilding

The Compendium stores your characters, locations, lore, and more.
You can:

  • write entries manually
  • or let AI help generate them

5. Explore Workshops

Creative tools designed to beat writers block:

  • character workshop
  • plot generator
  • setting creation
  • prose enhancer
  • dialogue helper

6. Everything Saves Automatically

Your work is stored locally using JSON.
No online accounts required.


🧭 Detailed Feature Breakdown

🗂 Project System

  • Multiple local projects
  • Autosave
  • Humanreadable JSON files
  • Quick switching
  • Clear internal structure (scenes, compendium, metadata)

🎬 Scene Editor

  • Clean writing interface
  • Scene titles & metadata
  • Draganddrop ordering
  • AI integration per scene
  • Automatic prompt construction using:
    • current scene
    • compendium
    • notes
    • style settings

📚 Compendium

  • Characters, places, species, lore, items
  • Tags and categories
  • Integrated into prompt builder

🧠 Workshop Chat

Chat with your AI about:

  • plot ideas
  • character outlines
  • worldbuilding prompts
  • style fixes
  • dialogue improvements
  • “retry / regenerate / variations” workflows

🧩 Prompt Builder / Codex

Automatically generates highquality prompts by combining:

  • scene text
  • compendium entries
  • metadata
  • tone/style guidelines

Ensures the AI responds consistently and in the desired format.


🔁 AI Integration

Supports:

  • local LLMs (LM Studio / Ollama / Jan)
    • Also has llama.cpp server integration, so you can just drop a model in the /models folder
  • OpenAI / OpenRouter

Features:

  • safe fallback prompts
  • retry logic
  • configurable token limits
  • userfriendly settings

📥 Writingway 1 Importer

Imports legacy Writingway 1 JSON projects:

  • scenes
  • compendium
  • project metadata
  • fixes ordering

Allows users to continue older stories seamlessly.


🖥 User Interface

Built with HTML + Alpine.js:

  • clean and accessible
  • minimal visual clutter
  • responsive design
  • intuitive sidebars
  • modals, tabs, and simple navigation

🔍 Update Checker

  • Notifies you if a new version is available
  • Checks GitHub releases
  • Optional and unobtrusive

⚙️ AI Setup (Optional)

Writingway 2 works out of the box with local models, meaning 100% privacy.

If using an online AI:

  1. Open Settings → AI Configuration
  2. Paste your API key
  3. Select a model

Writingway never sends your writing anywhere unless you configure it to do so.


Troubleshooting

Browser didnt open automatically

Visit: http://localhost:8000

start.bat closes immediately

Open a command prompt and run:

start.bat

to see the error message.

AI not responding

  • Make sure LM Studio / Ollama / Jan is running
  • Or check API key / endpoint URL

Scenes not saving

Ensure:

  • localStorage
  • cookies
    are enabled for localhost.

🛠 Development (For Contributors)

Requires Node 18+.

Install:

npm install

Run development server:

npm run dev

Project structure:

src/
  components/
  scenes/
  compendium/
  ai.js
  app.js
  workshop.js
  codex.js
  update-checker.js
  ...

Tests located in tests/.


📄 License

Writingway 2 is free and opensource software.
Released under the MIT License.


❤️ Thank You for Using Writingway 2

Writingway was built to help writers write — not struggle with tools.
If this app helps you bring your stories to life, it has succeeded.

Happy writing!

(And thanks for writing this readme, GPT-5.1)