Skip to content

About

A desktop AI assistant that turns on-screen context into answers, drafts, and rewrites—with your choice of model provider.

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Repository files navigation

ContextCue app icon

ContextCue

Ask about your screen. Turn ideas into drafts.

One floating assistant for understanding a page, drafting a reply, and finding the right words.

Download for macOS · Quick start · 简体中文

One shortcut · Your model provider · You decide what to send

ContextCue answers a question about the next steps in the visible team conversation.

Press a shortcut to open Ask AI over the window you’re using. Ask “What do I need to do here?” or say “Draft a polite decline, leaving the door open.” Read explanations directly; writing requests become drafts you can revise, copy, or insert. A separate shortcut takes you straight to writing suggestions.

What you can do

When you need to… ContextCue helps you…
Understand a page Ask for a summary, explanation, or next steps in a compact AI conversation, with optional page context.
Turn intent into a draft Say what you want to communicate in Ask AI. Writing requests open reply candidates you can revise, copy, or insert.
Find the right tone Revise a selected draft while keeping it in view. Compare alternatives, then return to the same Ask AI conversation.
Write beyond chat Draft text, rewrite a passage, complete a field, or compose a search query.
Use your own model Connect an image-capable model through a Responses or Chat Completions endpoint; keep multiple configurations.
Stay in control Review before copying or inserting. See provider-reported token usage locally. Nothing is sent or submitted automatically.

Works from app and browser windows without a per-app integration. On macOS, recognized text fields support insertion; copying is available when insertion is not supported. Search assistance drafts queries; it does not browse the web.

Quick start

1. Install or run locally

macOS: Open Releases, download the DMG for Apple Silicon (arm64) or Intel (x64), and move ContextCue to Applications. Check the release notes for signing and first-launch instructions; early-access builds may not be notarized.

From source: Requires Node.js 22.12+ and npm 10+.

git clone https://github.com/jastfkjg/ContextCue.git
cd ContextCue
npm install
npm run dev

macOS is the primary verified platform. Windows and Linux can run from source; Linux capture requires X11 and xdotool. Cross-app insertion is currently macOS-only.

2. Connect a model

The Setup guide walks you through adding a model and API key, verifying image input, checking screen access, and asking about or drafting a reply to a fictional conversation. macOS Accessibility permission is optional and enables insertion into supported fields.

Manage connections in Settings → Models; complete the required fields and changes save automatically. Use a model that supports image input for suggestions and page-aware questions. Text-only models work for questions, drafts, and subsequent draft revisions in Ask AI when page context is off. Model requests, including setup verification, may incur provider charges. Provider and environment configuration →

Check screen recording and optional insertion access in Settings → Permissions. Expand Test window capture, start the test, and switch to your target window within three seconds. Return to Settings to inspect the local preview.

3. Open a window and call ContextCue

Action macOS Windows / Linux
Open Ask AI (main entry) ⌘ ⇧ Space Ctrl ⇧ Space
Generate writing suggestions directly ⌘ ⇧ Enter Ctrl ⇧ Enter
Send an Ask AI question / add a newline Enter / Shift Enter Enter / Shift Enter
Switch suggestions ← / → ← / →
Apply the selected suggestion¹ Enter Enter
Submit revision instructions ⌘ Enter Ctrl Enter
Close the composer or panel Esc Esc

¹ Outside text inputs, with the revision composer closed. Inserts into a recognized macOS field, otherwise copies. It never sends the message. These are defaults for new installs; upgrades preserve existing shortcuts. Change shortcuts, writing language, and candidate count in Settings → General.

See it in action

From a question to a usable draft

Ask “What do I need to do before the review?” to understand the page. Then say “Draft a friendly reply confirming the time and promising to send the deck.” ContextCue opens writing candidates in the same floating window. Choose one, revise it, or return to Ask AI to continue the conversation. Open draft brings you back to the latest draft.

ContextCue offers reply candidates you can revise, copy, or insert.

Use the page chip to toggle screenshot context, or the refresh button to capture the original window again and start a new conversation. Refresh clears old questions and drafts only after capture succeeds. Switching to another app temporarily hides the panel; returning to the original window restores your work. Session and refresh details →

Make a draft sound like you

Choose Revise, describe the change, and get fresh alternatives in the same panel. Start with Shorter, Warmer, or More direct, or write your own instruction. The draft and revision instructions scroll separately, so you can keep the text in view. Collapse folds the instructions away; the top-right × closes the window. You can switch back to the original suggestions.

The real ContextCue revision composer with an English draft and the instruction: Make it shorter and friendlier. Keep the meeting time.

Set up your workspace

Home shows Ask AI, Quick writing, and model and screen-access status. Settings groups the controls into four tabs:

Tab What you can change
General Record shortcuts, choose a writing language, and set the number of suggestions.
Models Add model connections, test an endpoint, and choose the default.
Permissions Check screen and insertion access; test capture and inspect available windows.
About Check for updates and download a new version.

ContextCue’s General settings, showing the four settings tabs, Ask AI and Quick writing shortcuts, and writing preferences.

Settings save automatically. Writing language controls generated text; Match context follows the language of the page. Token usage in the sidebar shows usage by model, daily trends, and recent requests.

Your context, your choice

  • Capture on invocation. ContextCue does not record your screen in the background. A new invocation or explicit refresh takes a new snapshot.
  • Your chosen provider. Screenshots and relevant text are sent to the model endpoint you configure when used in a request. “Local-first” describes storage; inference may run remotely.
  • Local settings and history. Settings, memory documents, and usage records are stored on your device. Relevant enabled notes can be shared with your model for writing and personalized questions; past accepted replies are never automatically reused. Ask AI has an independent Memory switch, and results show the notes shared. API keys are encrypted with OS-backed storage; the entire data file is not encrypted.
  • You approve the words. Copy or insert explicitly. ContextCue never submits a form or sends a message for you. Screenshots are not automatically redacted.

Privacy details · Platform and capture limitations

Development and documentation

Built with Electron, React, and TypeScript.

npm run lint     # TypeScript validation
npm test         # Unit tests
npm run build    # Build the desktop app
npm run package  # Create an unpacked application
Explore Details
User guide Setup, model providers, local memory, token usage, architecture, and all development commands
Interaction flows Session boundaries, revisions, and acceptance checks
Updates and releases Packaging, signing, and in-app updates
Roadmap Current priorities and planned features

ContextCue is an early desktop MVP. Found something that could work better? Open an issue.

License

ContextCue is licensed under the MIT License. Commercial use, modification, and redistribution are permitted, including in closed-source products, provided the copyright and permission notices are preserved. The software is provided without warranty.

Third-party dependencies and assets remain subject to their respective licenses, including the Inter font (SIL Open Font License 1.1).

About

A desktop AI assistant that turns on-screen context into answers, drafts, and rewrites—with your choice of model provider.

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages