Dreamscribe

Dreamscribe is a multi-model storytelling studio with a cinematic landing page, local auth, collaborative AI writing tools, and a shareable library. Generate scenes with OpenAI, Anthropic, Gemini, or DeepSeek, polish drafts with targeted feedback, and publish your favourites to the community shelf.

Quick start

npm install
npm run dev

Scripts:

• npm run dev – start Vite dev server on http://localhost:5173
• npm run build – type-check and create a production build
• npm run preview – preview the production bundle locally

Note: Dependencies are declared but not yet installed. Run npm install from the project root before starting the dev server.

Bring your own API keys

Dreamscribe keeps secrets in the browser. Before you can generate text or request feedback:

1. Open the Studio (/studio) and choose an AI provider.
2. Paste your API key (OpenAI, Anthropic, Gemini, or DeepSeek) into the provider key field.
3. Keys are cached in localStorage only—Dreamscribe never transmits them to a server.

Provider	Endpoint	Default model
OpenAI	POST https://api.openai.com/v1/chat/completions	gpt-4o-mini
Anthropic	POST https://api.anthropic.com/v1/messages	claude-3-5-haiku-latest
Google Gemini	POST https://generativelanguage.googleapis.com/v1beta/models/{model}:generateContent	gemini-1.5-flash
DeepSeek	POST https://api.deepseek.com/chat/completions	deepseek-chat

Custom providers can be registered in the UI, but no connector ships by default for security reasons.

Features

• Immersive landing experience – Orbiting gradient animations, staged hero text, feature grid, workflow timeline, animated pricing, FAQ, and footer credit (“Made by Reno and Mason”).
• Local auth & security – Hashed credentials, persistent sessions, guarded studio route, and local-only key storage.
• Story studio
  • Prompt builder with tone/genre/POV controls and per-provider model selection.
  • Expanded model menus for OpenAI, Anthropic, Gemini, and DeepSeek with quick switching.
  • Multi-AI generation, dialogue/structure feedback passes, and inline debug console logging.
  • Publish workflow with summary/tags so you can share drafts to the community shelf.
• Achievements & collections – Unlock badges as you publish and review, and organise favourites into personal shelves.
• Public library – /stories showcases published tales with animated cards, average ratings, and quick navigation.
• Story detail view – Full prose display, quick-reference story metadata, collection controls, star reviews, and community comment log.
• Debug console – Toggle via navbar or Alt + D; filter requests/responses/errors and copy payloads for troubleshooting.

Project structure

+-- docs/
¦   +-- architecture.md
+-- public/
¦   +-- favicon.svg
+-- src/
¦   +-- components/
¦   ¦   +-- debug/
¦   ¦   +-- library/
¦   ¦   +-- layout/
¦   ¦   +-- navigation/
¦   ¦   +-- ui/
¦   ¦   +-- workspace/
¦   +-- context/
¦   +-- hooks/
¦   +-- lib/
¦   ¦   +-- clients/
¦   ¦   +-- prompts/
¦   +-- pages/
¦   ¦   +-- Auth/
¦   ¦   +-- Landing/
¦   ¦   +-- Library/
¦   ¦   +-- Workspace/
¦   +-- styles/
¦   +-- utils/
+-- index.html
+-- package.json
+-- vite.config.ts

Refer to docs/architecture.md for a deeper breakdown of flows and module responsibilities.

Debugging & instrumentation

• Every generation, feedback, and publish request logs to the in-app console with sanitized payloads.
• Use the “Debug” button in the navigation bar or press Alt + D to toggle the overlay.
• Filter events (info/request/response/error), inspect payloads, copy JSON, or clear the feed.

Roadmap suggestions

• Sync published stories and reviews to a hosted backend for multi-device access.
• Stream AI tokens in real time with cost telemetry and smarter retry heuristics for 429s.
• Add collaborative editing, inline comments, and reader shelving/bookmarks.
• Extend custom provider support with user-supplied headers/bodies and validation playgrounds.