Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
What This Skill Does
Downloads CBS Evening News and BBC News at Ten from YouTube on a daily schedule, extracts auto-captions, and uses DeepSeek to fix speech recognition errors in the subtitle files. Videos are served locally via HTTP with embedded web players on configurable ports.
Combines download, subtitle extraction, AI proofreading, and local HTTP serving into one automated pipeline instead of requiring separate tools for each step.
When to use it
- Watching last night's CBS or BBC news offline
- Correcting auto-generated subtitles for news broadcasts
- Running a local news video archive with a web player
- Scheduling automated nightly news downloads via cron
- Reviewing AI-corrected subtitle diffs to catch transcription errors
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: Download today's CBS and BBC news with proofread subtitles
- 1Run video_download.py --cbs --bbc --proofread to fetch both broadcasts from YouTube
- 2yt-dlp extracts video and auto-caption VTT files into cbs-live-local/ and bbc-news-live/
- 3subtitle_proofreader.py sends each VTT to DeepSeek, which corrects recognition errors
- 4Corrected text and a corrections diff are saved alongside the original backup
- 5setup_server.sh starts HTTP servers on ports 8093 and 8095 with embedded players
Both news videos available at http://IP:8093 and http://IP:8095 with corrected subtitles
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.