live-stream-monitor

Monitor live streams (YouTube, Bilibili) and get notified when specific keywords are mentioned. Uses browser SpeechRecognition API for real-time transcriptio...

Security Scan

The skill's code and description mostly match its stated purpose, but there are important mismatches (missing install/dependency declarations and an automated browser that auto-grants microphone access) that a user should understand before installing.

Purpose & Capability

Name and description (live stream monitoring via browser SpeechRecognition) align with the included monitor.js which uses Playwright to open a headed Chromium and drive the Web Speech API. However, the SKILL metadata and SKILL.md omit that this is a Node/Playwright script requiring Playwright and Node runtime — a capability mismatch between 'instruction-only' metadata and the shipped code.

Instruction Scope

SKILL.md instructs monitoring live streams and requiring microphone/SpeechRecognition only. The runtime code stays within that scope (opens stream pages, runs in-browser SpeechRecognition, prints transcripts). It does not read unrelated files, exfiltrate data, or contact third‑party endpoints in the code.

Install Mechanism

There is no install spec even though monitor.js requires the 'playwright' Node package and a Node runtime. The skill will not run as-is in most agents/environments without installing Playwright/Node manually. Missing dependency declaration is a practical risk and a registry metadata inconsistency.

Credentials

No environment variables, credentials, or config paths are requested. The code does request microphone access in the browser (appropriate for transcription) and uses the Chromium flag '--use-fake-ui-for-media-stream' to bypass permission prompts — this is functional for automation but affects privacy/consent behavior.

Persistence & Privilege

The skill is not always-enabled and does not request persistent platform privileges. It does require running a headed browser instance (visible UI) and will prompt or auto-grant microphone access; it does not attempt to modify other skills or system-wide agent settings.

What to consider before installing

This skill appears to do what it says (open a browser and run in-browser speech recognition), but there are some practical and privacy considerations before you install or run it: - Dependency gap: monitor.js requires Node and the Playwright package, but the skill metadata provides no install instructions. You will need to install Node and run npm install playwright (or the appropriate Playwright bundle) before the script will work. - Headed browser & microphone: The script launches Chromium in headed mode and uses the '--use-fake-ui-for-media-stream' flag so the page can access the microphone without a prompt. That makes automation easier but means the local microphone can be captured by the opened page; only run this on a trusted machine and network. - No network exfiltration in code: The script logs transcripts to the console and does not POST data to remote servers, but you should manually review the code yourself before running. If you need stronger guarantees, run it in an isolated VM or sandbox and restrict outbound network access. - If you expect a pure 'browser extension' or in-browser user workflow, note this is a Node+Playwright automation script (not a browser extension). If you want an agent to run it, ensure the runtime environment can provide a headed browser and the required Node dependencies. If you want me to, I can: (1) produce exact npm install commands and a minimal package.json for this script; (2) walk through how to run it safely in a sandbox/VM; or (3) review any modifications if you want transcripts saved to a file or posted to a secure endpoint.

Monitor live streams and get notified when keywords are mentioned!

What It Does

• Monitor live streams in real-time
• Alert when your keywords are mentioned
• Speech-to-text transcription via browser

Features

• ⭐ Keyword Alerts - Set keywords, get instant notifications
• Live Detection - Auto-detect stream start/end
• Speech Transcription - Real-time speech-to-text

Supported Platforms

• YouTube Live
• Bilibili Live

Usage

Monitor [keyword] on [stream URL]
Example: Monitor "discount" on https://youtube.com/watch?v=xxx

Requirements

• Browser with microphone access
• SpeechRecognition API support (Chrome/Edge)

Note

Uses browser's built-in SpeechRecognition API for transcription - no external APIs needed!