
from video-podcast-maker501
Automates end-to-end production of long-form video podcasts using TTS, Remotion, and FFmpeg—supports multi-language output and Bilibili/YouTube publishing workf
Automates an end-to-end pipeline that turns a topic into a production-ready video podcast. The skill handles research, scriptwriting, TTS audio generation, Remotion composition and studio preview, thumbnail generation, and final MP4 rendering with background music. It includes a design-learning subsystem that extracts visual patterns from reference videos or images and applies them as style profiles to new compositions. The workflow is optimized for horizontal Bilibili-style knowledge videos but can generate vertical highlights for shorts.
Use this skill when you want a coding agent to produce a complete video from a topic prompt ('Make a video podcast about X') with minimal manual steps, or when you need design-consistent templates learned from reference videos. It’s appropriate for: batch content production, repackaging articles into video, rapid prototyping of visual styles from references, and generating publish-ready 4K/1080p outputs.
Best fit for coding agents and Claude Code-style tooling that can run Node/Remotion, Python/FFmpeg, and TTS backends. Works with agents that can call shell tools, manage files, and launch Remotion Studio for user review.
Video Podcast Maker is a comprehensive end-to-end pipeline for producing 4K video podcasts from a topic, covering research, scripting, TTS, Remotion composition, rendering, and Bilibili/YouTube publishing. It features 13 scripts with a CLI dispatcher, standardized JSON output envelope, and multi-backend TTS support (Edge/Azure/Doubao/ElevenLabs/OpenAI/Google). Scripts are well-written with proper error handling but depend on local module siblings (tts/, cli_envelope.py) that can't be resolved when run in isolation. No security concerns — no hardcoded credentials, no destructive commands, and the update check is consent-based.
tts (local module package)cli_envelope (local module, needed by most scripts)One of the most polished and comprehensive skills audited. Production-quality CLI envelope with error codes, request IDs, and latency tracking. The skill follows the AgentSkill spec excellently with progressive disclosure via references/, dependency declaration, and a clear 15-step workflow. The only notable gap is that most scripts can't run in isolation due to local module dependencies, but this is expected behavior for a skill designed to operate within its directory context.