openjutsu
영상에서 원클릭으로 인물 교체, ComfyUI·MiniMax H3 기반 오픈소스 로컬 힉스필드 겐주츠 대안, FLUX.3 Klein 재조명 지원
저장소 정보는 아직 수집하지 않았습니다.
아직 감지된 업데이트가 없습니다. GitHub 릴리스가 나오면 여기에 보입니다.
메이커가 쓴 소개는 위의 한 줄이 전부입니다. 더 자세한 내용은 저장소 README에서 읽을 수 있습니다. 저장소 README ↗
README 전체를 불러오는 중입니다.
README 전문 보기 ↗Publish the article. Let the ghost writer handle LinkedIn.
A publishing platform with an AI ghost writer built into the editor — it reads your finished article and writes the LinkedIn post for you, hook and all, ready to send from the same page.
Ghost writer · Write · Teams · Comments · Search · Read · Try it
Most publishing tools give you a text box and a Publish button. The hard parts happen everywhere else: the draft you paste into a chat for feedback, the comment thread that loses track of which paragraph it was about, the LinkedIn post you rewrite by hand an hour after publishing.
Quill puts all of it in one place.
A clean, focused composer with everything a post actually needs — headings, emphasis, highlights, alignment, code blocks, and images you drop straight into the page. Add a cover image and a summary and your post looks like a real publication, not a text dump.
Save drafts as long as you like. Nothing goes live until you say so.
Feedback shouldn't mean publishing early or emailing a Google Doc.
Create a team from people who already write on Quill, then share a post into it right from the editor — to one team or several. Unpublished drafts become readable by those members and nobody else. Your teams page collects both sides of it: what you've shared out, and what's been shared with you, each post one click from a full read.
When the draft is ready, publish it. The team keeps its copy; the world gets the published one.
Margin notes. Readers highlight the exact passage they're responding to and leave a note anchored to it — not stranded at the bottom of the page. Nothing appears on your post until you approve it. You review each note from your profile and approve or reject it, so the conversation lives in the margins and the page stays yours.
Comments at the foot of the post. For everything that isn't about one specific line — a reaction, a question, a thank-you. These go up as soon as they're written.
Either way you hear about it immediately: Quill holds a live connection while you're on the site, so a new note, a reply, or an approval lands in your notifications the moment it happens, with no refresh and no polling delay.
Finished a post? Quill reads it and writes the LinkedIn version for you — a real hook, the concrete details, the right length for a feed.
Edit it, regenerate it if the angle's wrong, then publish it to your profile without leaving the page. The draft is saved with your post, so you can come back to it.
Search the whole library by story title or by author, straight from the stories page. Results narrow as you type, and the query lives in the URL — so a search is something you can bookmark, share, or step back through with the back button.
No feeds to fight, no popups, no paywall prompts. Typography built for long-form reading, a sane line length, and a light and dark theme that are both worth using. Every writer gets a profile page that collects their published work in one place.
Quill is a Next.js 16 app on React 19 talking to a Hono API running on Cloudflare Workers, with PostgreSQL behind Prisma Accelerate. Social drafts are generated by Workers AI, and live notifications ride a WebSocket backed by a Durable Object per user. It runs at the edge, so it's fast wherever you are.
Security is not an afterthought: passwords are hashed with PBKDF2-HMAC-SHA256, OAuth tokens are encrypted at rest with AES-256-GCM, and post content is sanitized both on write and on render.
You'll need Node 20+, pnpm 9+, a Postgres database, and a Cloudflare account.
git clone <your-repo-url> quill
cd quill
pnpm install
pnpm dev
Frontend at http://localhost:3000, API at http://localhost:8787.
Full setup — environment variables, OAuth apps, Cloudinary, migrations, deployment — is covered in the developer documentation, along with the architecture diagram, data model, and API reference.
Email and password, or one click with GitHub, Google, or LinkedIn.
MIT — do what you love with it.