content-curation-and-feeds

# Content Curation & Feed Monitoring This skill covers tracking, monitoring, and reformatting incoming public content feeds (blogs, RSS/Atom feeds) and public video transcripts (YouTube) to build summaries, knowledge base notes, or social media threads. --- ## 1. Blog & RSS Monitoring (blogwatcher) Track blog and RSS/Atom feed updates using the `blogwatcher-cli` tool. Supports feed auto-discovery, HTML scraping fallback, OPML imports, and read/unread article management. ### Prerequisites `blogwatcher-cli` installed. Database path defaults to `~/.blogwatcher-cli/blogwatcher-cli.db` (persistent Docker volume: `-v blogwatcher-cli:/data -e BLOGWATCHER_DB=/data/blogwatcher-cli.db`). ### Common Commands - **Add Blog:** `blogwatcher-cli add "My Blog" https://example.com` - **List Tracked Blogs:** `blogwatcher-cli blogs` - **Scan Feeds:** `blogwatcher-cli scan` - **List Unread Articles:** `blogwatcher-cli articles` - **Mark Article Read:** `blogwatcher-cli read <ID>` - **Mark All Read:** `blogwatcher-cli read-all` --- ## 2. YouTube Content Parsing (youtube-content) Extract transcripts from YouTube videos (standard URLs, shorts, embeds, or raw 11-char IDs) and reformat them into structured summaries, chapters, blog posts, or social media threads. ### Prerequisites ```bash pip install youtube-transcript-api ``` ### CLI Transcript Extraction - **JSON with metadata:** `python3 ~/.hermes/skills/media/content-curation-and-feeds/scripts/fetch_transcript.py "URL"` - **Plain text only:** `python3 ~/.hermes/skills/media/content-curation-and-feeds/scripts/fetch_transcript.py "URL" --text-only` - **With timestamps:** `python3 ~/.hermes/skills/media/content-curation-and-feeds/scripts/fetch_transcript.py "URL" --timestamps` ### Synthesis & Output Formats After fetching a transcript, transform it into one of these desired shapes: - **Chapters:** Chronological list of topics with timestamps. - **Summary:** Concise 5-10 sentence overview of the video. - **Twitter Thread:** Numbered posts (each under 280 characters). - **Blog Post:** Clean markdown article with headers, sections, and takeaways. --- ## Guidelines & Error Handling - **Large Transcripts:** If a video transcript exceeds 50K characters, split it into overlapping chunks (~40K chars, 2K overlap), summarize each chunk, and synthesize the summaries. - **Scanned/No Transcript:** If transcripts are disabled or unavailable, notify the user. - **Database Backups:** Keep the blogwatcher SQLite database backed up regularly if monitoring high-value feeds.