Video Archive Workflow

Video download, metadata extraction, deduplication, subtitle handling, and archive organization for URLs from YouTube/Shorts, Xiaohongshu, Bilibili, and X/Tw...

This skill's instructions clearly expect yt-dlp (for JSON metadata) and a tool for lossless container conversion (e.g., ffmpeg), plus the ability to download files and write archives to disk — but the metadata declares no required binaries, no install steps, and no archive destination. Before installing or enabling this skill, ask the publisher to: (1) explicitly list required binaries (yt-dlp, ffmpeg or equivalent) and their trusted install sources; (2) declare where archive files will be written (a configured archive root path or env var) and what filesystem permissions are needed; (3) clarify whether network access is required and what endpoints will be contacted; and (4) confirm that the agent will prompt before downloading/writing files (the SKILL.md says to download only when asked — verify this behavior). If you cannot get those clarifications, treat the skill as risky: ensure the agent runs in a sandbox with limited filesystem and network permissions, or do not enable autonomous invocation for this skill.

Type: OpenClaw Skill Name: video-archive-workflow Version: 1.0.0 The skill bundle describes a standard video archiving workflow using yt-dlp to extract metadata and download content from platforms like YouTube, Bilibili, and X/Twitter. The instructions in SKILL.md and references/field-schema.md focus on data normalization, deduplication, and file organization, including safety checks that require user confirmation for ambiguous cases or file conflicts.