QUILL for All is an open-source, accessibility-focused editor from Community Access.

QUILL for All helps people write, edit, convert, compare, and publish documents in a screen-reader-friendly environment.

QUILL stands for Quality, Usable, Inclusive, Lightweight, Literate.

QUILL for All: A quality, usable, inclusive, lightweight, and literate editor built for everyone who writes, codes, learns, and creates.

Expand your editor's powers with the Quillin Hub. Discover a curated gallery of community-created extensions -- from research tools to accessibility auditors -- all verified for security and WCAG 2.2 AA compliance.

Quill is a screen-reader-first writing and document environment for Windows and macOS, focused on practical keyboard workflows, stable editing, and accessible diagnostics/support flows.

Quill is designed to stay focused and useful:

• Quality -- dependable, polished, and serious enough for real work.
• Usable -- built around practical keyboard, screen reader, and low-friction editing needs.
• Inclusive -- designed from the beginning for blind users, screen reader users, keyboard users, and people with different skill levels.
• Lightweight -- fast, focused, not bloated, and friendly to people who just want to write or edit.
• Literate -- about words, code, Markdown, documents, learning, and thoughtful communication.

Current release line: 0.7.0

Highlights since 0.5.0 include:

• Insert menu with searchable Markdown/HTML insertion.
• Word Prediction with Ctrl+. plus HTML/Markdown tag IntelliSense.
• New snippet system with Ctrl+Shift+Grave, S insertion, trigger expansion, and starter packs.
• Release-safety default: Word and CSV open in the normal plain-text editor surface.
• Structured Word view and CSV grid code paths remain in-repo behind an internal gate for continued verification.
• Expanded structured intake for .doc/.docx, .ppt/.pptx, .xlsx/.xls, .pages, and low-confidence PDF fallback via MarkItDown when available.
• Writing Assistant shell with prompt presets, generated tool suggestions, and a restricted Python runner (import allowlist, resource limits, confirmation required — not a security boundary; see the user guide).
• AI Connection workflow from both Preferences and the AI menu, with provider-aware host defaults.
• Verify Connection, List Models, and Recommend Model actions in AI Connection settings.
• AI menu status line with plain-language detail (Ready or Needs attention) and immediate accessible feedback.
• BITS Whisperer rollout surfaces for provider onboarding, readiness checks, status-page live updates, and guarded download queue controls.
• General Preferences controls for AI enable state, BW Safe Mode Lock, auto-open status behavior, and refresh cadence.
• Optional Ollama cloud key mode over HTTPS (no local Ollama required for cloud endpoint access).
• In-App Preview and Side-by-Side Preview with a dedicated Focus Preview command.
• Heading styling tools to apply font family, size, and alignment to current-level or all headings in Markdown/HTML.
• Heading Organizer (Ctrl+Shift+Grave, O — shown to users as QUILL Key + O) with keyboard-driven heading level changes, section reordering, and accessibility validation.
• QUILL Quick Nav mode (browse-style cursor navigation) activated with the QUILL key (Ctrl+Shift+Grave — shown to users as QUILL Key), with mnemonic single-key movement for links, lists, list items, tables, block quotes, bookmarks, code blocks, table of contents, headings, heading levels (1 through 6), paragraphs, sentences, and blocks.
• Watch Folder automation under Tools -> Dictation to auto-open newly dropped supported files.
• Unified first-run Personalise QUILL wizard (re-runnable from Help -> Personalise QUILL) covering keyboard pack, feature profile, remote access, AI, reading/accessibility, writing tools, and startup behaviour.
• Search menu simplification with replace-all inside the Replace dialog.
• Unified diagnostics-backed support flow under Help -> Report a Bug.
• Menu IA refinement, including Insert before View and Search after View.
• Documentation refresh with regenerated Markdown/HTML/EPUB artifacts.

AI quick start:

1. Open AI -> AI Connection....
2. Choose provider (Ollama (local), Ollama Cloud (API key), or Custom HTTP).
3. Enter host/model/key as needed.
4. Save settings; Quill auto-verifies and updates the AI status line.
5. Use List Models to choose a provider-returned model.

Ollama Cloud onboarding is available in this same flow. If you have an API key, Ollama Cloud offers a free personal-use tier with lower usage limits.

Snippet workflow quick start:

1. Press Ctrl+Shift+Grave, S to open Insert Snippet.
2. Type to filter by snippet name, trigger, or body text.
3. Use arrow keys to choose, press Enter to insert, and fill placeholders when prompted.

Related commands:

• Ctrl+.: Word Prediction (words, HTML tags, Markdown tags).
• Ctrl+Shift+Grave, Shift+S: Manage snippets (create, edit, delete, import/export, starter packs).
• Preferences -> Install Starter Snippet Packs: install sample packs for writing, developer flow, and accessibility/support notes.

Watch Folder quick start:

1. Open Preferences -> Watch Folder Automation.
2. Choose a folder where you will drop supported Quill files.
3. Enable watch folder monitoring.
4. Turn on auto-start if you want it running every launch.
5. Drop supported files into the folder; Quill opens them automatically.

QUILL Quick Nav quick start:

1. Press Ctrl+Shift+Grave to enter QUILL Quick Nav mode.
2. Press H for next heading, or Shift+H for previous heading.
3. Press 1 through 6 for next heading at that level, or Shift+1 through Shift+6 for previous heading at that level.
4. Press A for links, L for lists, I for list items, T for tables, Q for block quotes, B for bookmarks, and ' for code blocks.
5. Press C to open table of contents, P for paragraphs, S for sentences, and Tab for blocks. Use Shift with any movement key to reverse direction.
6. Press ] to jump to the first line after the current list or table. Press [ to jump to the line above it.
7. In Preferences -> General, use Preload QUILL browse cache in background. It is on by default; if off, Quill builds the cache the first time you use Quick Nav.
8. Press Esc to leave QUILL Quick Nav mode.

How tracking works for Markdown and HTML:

• Quill builds a per-document navigation index in memory and reuses it while the text and markup type stay unchanged.
• Headings are indexed from parsed Markdown and HTML heading structures.
• List-item anchors are indexed from Markdown list syntax and from HTML <li> tags.
• Paragraph anchors are indexed by blank-line paragraph boundaries for text/Markdown and by block-level tags in HTML (p, li, blockquote, pre, h1 to h6, td, th).
• Sentence anchors are indexed from sentence-ending punctuation boundaries.
• The index is invalidated on document edits, full-text replacements, and tab switches.
• Quick Nav movement is cursor-only and non-editing by design.

• Cross-platform. Quill runs on Windows and macOS from one codebase. The macOS build ships as a signed, notarized .app; screen-reader announcements route to VoiceOver on macOS and to NVDA/JAWS/Narrator (via Prism) on Windows.
• Ask Quill chat. An on-device AI chat (AI -> Ask Quill Chat) rendered as a fully accessible WebView document: each turn is a heading you can navigate, new replies are announced, the message box lives in-page, and Escape closes it. Verified in NVDA, JAWS, and VoiceOver.
• On-device AI, no cloud required. macOS uses Apple Foundation Models (Apple Intelligence); Windows/Linux use llama.cpp (CPU, GGUF). You can optionally connect Ollama (local), Ollama Cloud (API key), or a custom HTTP endpoint. The assistant defaults to answering in chat and never edits your document without approval.
• Accessible WebView, preview, and dialogs are built on the open-source wx-accessible-webview library (extracted from Quill), which also powers the live Markdown/HTML preview, the About dialog, and the update/consent dialogs.
• Train Writing Style (AI -> Train Writing Style...) conditions the assistant on your own writing.

• quill/ -- application code.
  • quill/core/ -- core logic and document operations.
  • quill/ui/ -- wxPython interface and dialogs.
  • quill/platform/windows/ -- Windows integration points.
  • quill/platform/macos/ -- macOS integration (VoiceOver announcements, Keychain, high-contrast, screen-reader detection).
  • quill/core/ai/ -- on-device assistant backends (Apple Foundation Models, llama.cpp) and the Ask Quill agent.
• docs/ -- product docs and generated artifacts.
  • QUILL-PRD.md (+ .html, .epub)
  • userguide.md (+ .html, .epub)
  • announcement.md (+ .html, .epub) -- the 0.5.0 product announcement
  • engineering.md, planning.md, accessibility.md, features.md, qa.md -- consolidated reference docs (each + .html, .epub)
• tests/ -- unit/integration/accessibility/performance tests.
• scripts/ -- release, validation, and maintenance helpers.

Runs on Windows and macOS (Python 3.12).

1. Install Python 3.12.
2. Install dependencies:
   • pip install -e ".[ui,dev]"
   • On-device AI for Ask Quill: add ai on Windows/Linux (pip install -e ".[ui,ai]", pulls llama.cpp). macOS uses Apple Foundation Models — no extra needed.
   • Tip: llama-cpp-python has prebuilt CPU wheels but no Windows wheel on PyPI itself; if pip starts a slow source build, force the wheel with pip install --only-binary=llama-cpp-python ... (the project's extra index already points at the CPU wheels).
   • Offline speech (Tools > Speech > Whisperer): no Python package is required — the whisper.cpp engine ships with the Windows installer (the offline speech engine component) or you can drop it under tools/speech/whispercpp. To transcribe non-WAV audio/video (mp3, m4a, mp4, mov, ...), install ffmpeg and put it on your PATH (for example winget install Gyan.FFmpeg); QUILL invokes a system/user-installed ffmpeg and does not bundle it (ffmpeg is GPL/LGPL). For the optional GPU-accelerated engine, add pip install faster-whisper.
3. Launch:
   • python -m quill (on Windows, pythonw -m quill for no console window)

To build a signed, notarized macOS app, see scripts/build_macos.sh and docs/QUILL-PRD.md.

Optional launch flags:

• --safe-mode
• --reset-profile
• --version (prints version and exits without launching UI)
• --line N --column M (place cursor for the first startup file)
• --new-window (force a new process instead of forwarding to existing instance)
• --wait (when forwarding to an existing instance, wait until it exits)
• --diagnostics
• --dump-stacks

Examples:

• python -m quill --version
• python -m quill notes.md --line 120 --column 1
• python -m quill --new-window notes.md

• Lint: ruff check .
• Tests: pytest tests/unit/ tests/stability/ -q

Generate docs artifacts from docs/*.md with Pandoc:

• HTML: pandoc docs\\<name>.md -f gfm -t html5 -s -o docs\\<name>.html
• EPUB: pandoc docs\\<name>.md -f gfm -t epub3 -o docs\\<name>.epub

Artifact parity guard:

• python scripts/check_docs_artifacts.py

This fails when a docs/*.md source changed but matching .html/.epub files were not updated. The installer ships the user-facing guides, while the GitHub Pages docs hub exposes the PRD and engineering docs for anyone who wants the deeper implementation detail.

Run the full release readiness flow:

• python scripts/release_readiness.py

This runs:

1. Lint (ruff check .)
2. Dependency audit (pip-audit --strict)
3. Tests (pytest -q)
4. Docs rebuild for docs/*.md (HTML + EPUB)
5. Docs artifact parity check
6. Release corpus verification

• CI workflow: .github/workflows/accessibility-ci.yml
• Security workflow: .github/workflows/security-ci.yml
• Windows release workflow: .github/workflows/windows-release.yml
• Docs site workflow: .github/workflows/github-pages.yml (publishes the docs hub and updates feed)

Use Help -> Report a Bug inside Quill. The flow supports diagnostics bundle generation, report preview, clipboard copy, and browser handoff to support submission.

Community contributions are welcome.

• Read CONTRIBUTING.md for setup, workflow, and PR expectations.
• Read CODE_OF_CONDUCT.md before participating.
• Read SECURITY.md for private vulnerability reporting.
• Read PRIVACY.md for data handling and retention behavior.
• Read RESPONSIBLE_AI_USE.md for ethical and accountable AI use requirements.
• Read GOVERNANCE.md for project decision model.
• Read MAINTAINERS.md for maintainer responsibilities.
• Contributor graph: contrib.rocks / QUILL

• Use GitHub Discussions for Q&A and feature ideation.
• Use GitHub Issues for confirmed bugs and scoped feature requests.

• Release process and branch policy: RELEASE.md
• Security advisory runbook: docs/QUILL-PRD.md

MIT. See LICENSE.

QUILL for All is an independent open-source project by Community Access. It is not affiliated with, sponsored by, or endorsed by Quill.js, QuillBot, Quill.org, or any other similarly named product, project, company, or organisation.

All trademarks are the property of their respective owners.

See TRADEMARKS.md, NOTICE, and THIRD_PARTY_NOTICES.md for more information.