OmniKey-AI

mcp
Guvenlik Denetimi
Basarisiz
Health Gecti
  • License — License: MIT
  • Description — Repository has a description
  • Active repo — Last push 0 days ago
  • Community trust — 14 GitHub stars
Code Basarisiz
  • rm -rf — Recursive force deletion command in .github/workflows/release.yml
  • rm -rf — Recursive force deletion command in api/package.json
  • network request — Outbound network request in api/package.json
  • network request — Outbound network request in api/public/app.js
Permissions Gecti
  • Permissions — No dangerous permissions requested

Bu listing icin henuz AI raporu yok.

SUMMARY

OmniKey AI is a productivity tool that enhances your workflow with AI-powered prompt enhancements, grammar correction, and custom tasks.

README.md

OmniKey AI Icon

OmniKey AI

Select text anywhere. Press a shortcut. AI rewrites it in place.
Prompt enhancement, grammar fixes, and custom AI tasks in every app on macOS and Windows.

npm downloads npm version GitHub stars Latest release License: MIT Platforms

Website  ·  Docs  ·  Changelog  ·  Discussions

OmniKey AI demo: selecting rough text and pressing Command-E rewrites it into a clear prompt in place

Available on macOS (menu bar app) and Windows (system tray app).

🌐 Full documentation, downloads, and product details live at omnikeyai.ca. This README only covers the essentials.

Getting Started

  1. Install the CLI:

    # macOS (recommended)
    brew tap GurinderRawala/omnikey-ai https://github.com/GurinderRawala/OmniKey-AI.git
    brew install omnikey-cli
    
    # or via npm (cross-platform)
    npm install -g omnikey-cli
    
  2. Onboard and configure a provider — pick OpenAI, Anthropic, Google Gemini, or an OpenAI-compatible open model, and optionally a web search provider:

    omnikey onboard
    
  3. Start the persistent daemon:

    omnikey daemon
    
  4. Download the desktop app — macOS · Windows.

Features

A quick reference of the core capabilities and the commands or shortcuts that drive them. Full details live on the website.

  • Prompt Enhancement — ⌘E / Ctrl+E rewrites the selected text into a stronger prompt.

  • Grammar & Clarity Fix — ⌘G / Ctrl+G cleans up grammar without changing meaning.

  • Custom Tasks — ⌘T / Ctrl+T applies your saved task instructions to the selected text.

  • Multiple LLM providers — switch between OpenAI, Anthropic, Google Gemini, and OpenAI-compatible open models during onboarding or via omnikey set.

  • Web Search — opt in during onboarding. Supports DuckDuckGo, Serper, Brave Search, Tavily, and SearXNG.

  • Authenticated Browser Sessions — omnikey grant-browser-access sets up a dedicated debug profile so the agent can read logged-in pages (Chrome, Brave, Edge, Arc, Vivaldi, Opera, Chromium, and Safari on macOS).

  • MCP Servers — the agent sees every configured server and connects only to the one it needs. Mention the configured MCP name in your task, such as “Use Slack to search messages”:

    omnikey mcp add
    omnikey mcp list
    omnikey mcp toggle <id>
    
  • Telegram Integration — run tasks and get notifications from any device:

    omnikey telegram start
    omnikey telegram status
    omnikey telegram logs
    omnikey telegram stop
    

    In Telegram, use /cmd to start or resume an agent task, /model to change the active provider's agent model, /task to fetch progress or the last result, and /stop to abort a running session.

  • Scheduled Jobs — automate recurring or one-time prompt runs from the desktop app or the CLI (omnikey schedule add / list / remove).

The @omniAgent

Select text starting with @omniAgent and press ⌘T / Ctrl+T to hand off the task to the autonomous agent. It can:

  • Run shell commands on your machine to inspect files, processes, environment, or run builds.
  • Read your live browser tabs (including authenticated ones) for context from internal dashboards, private docs, and paid tools — no copy-paste required.
  • Search the web via your configured provider (falls back to DuckDuckGo).
  • Use any MCP server you have registered (databases, APIs, custom tools).
  • Combine your saved task instructions automatically when relevant.

Use @omnikeyai (without the agent prefix) for plain Q&A in the context of your selected text. For the full agent design and tool list, see omnikeyai.ca.


Developers

OmniKey AI is a Yarn monorepo with three TypeScript workspaces (api, cli, telegram) plus native desktop clients in Swift (macOS) and C# (Windows). For setup, build/test workflows, and a tour of the main commands, see DEVELOPMENT.md.

License & Terms

OmniKey AI is open-source software released under the MIT License.
Use of any component is governed by the Terms and Conditions:

  • Desktop app (macOS menu bar / Windows tray) — the first launch shows a
    Terms & Conditions dialog you must accept before the app becomes usable.
    Acceptance is remembered per terms revision; a materially updated TERMS.md
    re-prompts on the next launch.
  • CLI (omnikey-cli) — the interactive omnikey onboard command handles
    provider configuration, not terms acceptance. Installing or running the CLI
    is treated as acceptance of the same Terms & Conditions.

The software is provided as-is, without warranty of any kind; OmniKey and
its contributors are not liable for any damages, data loss, third-party
charges, or issues arising from AI-generated content or from use of the
software.

Yorumlar (0)

Sonuc bulunamadi