daily-voice-quote_skill

Overview

This skill automates a daily voice-quote delivery that produces three outputs: a cloned-voice audio, a static-cover video, and an optional HeyGen digital-human video. It runs on a schedule, selects a quote, synthesizes the owner’s voice (or fallback TTS), generates a cover image, combines assets into an MP4, and sends them to the owner via the configured channel. The flow supports LINE-specific media formatting and optional HeyGen integration for higher-fidelity video avatars.

How this skill works

Each morning the skill picks a quote (holiday-aware rules), renders a voice file using ElevenLabs voice cloning or built-in TTS, and converts audio to the required M4A format. It generates a portrait-focused cover image from the owner’s photos, stitches the image and audio into an H.264/AAC MP4 static video, and optionally requests a HeyGen avatar video. Finally, it uploads assets to an HTTPS host and pushes audio/video messages to the configured channel (LINE, Telegram, Discord, etc.).

When to use it

• Daily motivation or reminder sent to a single user or group
• Distributing branded voice-based content using owner’s cloned voice
• Creating short vertical clips for mobile-first channels (9:16)
• Replacing manual quote posting with automated scheduled delivery
• Adding optional HeyGen videos where a trained avatar is available

Best practices

• Provide 1–3 high-quality, front-facing photos in avatars/ for reliable cover generation
• Use an ElevenLabs voice clone (1–3 minutes sample) for natural, consistent audio; fallback to built-in TTS if needed
• Ensure media hosting supports HTTPS and Range requests (LINE requires this)
• Convert audio to M4A (AAC) and video to MP4 (H.264 + AAC) to meet client playback rules
• Keep holiday and extreme-weather quote lists in references/ for predictable selection

Example use cases

• Send a personalized daily quote to a user’s LINE each morning with owner’s voice and a vertical cover video
• Create a daily motivational clip for a team channel on Discord or Telegram using built-in TTS when no voice clone exists
• Optionally generate a HeyGen avatar video for premium subscribers while still providing MP4 fallback
• Use cron to schedule a 6:00 AM Asia/Taipei job that pushes audio + video to a user ID
• Automate content for social stories by exporting the vertical MP4 and preview image for reposting