The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/scrape_cross_niche_outliers.py uses subprocess.run to execute the yt-dlp command-line utility for scraping YouTube video metadata.
Evidence: Found in scrape_keyword, scrape_channel, and get_channel_average functions using the subprocess.run() method with arguments passed as a list.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data (YouTube video titles and transcripts) by passing them directly into Claude API prompts for summarization and title variant generation without explicit boundary markers or sanitization instructions.
Ingestion points: Video titles and transcripts fetched from YouTube via yt-dlp, youtube-transcript-api, and Apify.
Boundary markers: None detected; variables are interpolated directly into f-string prompts in summarize_transcript and generate_title_variants within scripts/scrape_cross_niche_outliers.py and scripts/scrape_cross_niche_tubelab.py.
Capability inventory: The agent has the capability to write data to Google Sheets via gspread and the Google Sheets API.
Sanitization: No specific filtering or escaping of content found before interpolation into LLM prompts.
[EXTERNAL_DOWNLOADS]: The skill fetches content from external sources to perform its primary function.
Fetches video metadata from YouTube using yt-dlp.
Fetches video transcripts using youtube-transcript-api and the Apify karamelo/youtube-transcripts actor.

cross-niche-outliers