{"slug":"ytdlp-agent","title":"yt-dlp Agent — YouTube & Video CLI","tags":["ytdlp","youtube","download","transcript","cli","proxy"],"agent_summary":"Agent-native yt-dlp CLI wrapper with Oxylabs ISP proxy by default. Commands: download, audio, info, transcript, thumbnail, playlist, search, channel. JSON output for pipelines.","trigger_phrases":["yt-dlp agent","download youtube video","youtube transcript cli","ytdlp download","extract youtube audio","youtube playlist download","video info cli","ytdlp proxy"],"runnable":true,"markdown":"\n# yt-dlp Agent — YouTube & Video CLI\n\nAgent-native CLI wrapper for yt-dlp. Oxylabs ISP proxy by default. All commands return JSON for pipeline use.\n\n## Install\n\n```powershell\ncd D:\\ClaudeDev\\00_GITHUB\\cliit\\clis\\ytdlp-agent\npip install -e .\n```\n\n## Proxy Rule\n\nAll YouTube and video downloads must use the Oxylabs proxy (home IP rule). The `--no-proxy` flag is available for non-scraping one-off checks.\n\n## Commands\n\n### download — video or audio\n\n```bash\nytdlp-agent download # MP4, current dir\nytdlp-agent download -o D:/Downloads # custom dir\nytdlp-agent download --format mp3 # extract MP3\nytdlp-agent download --format wav # extract WAV\nytdlp-agent download --json # JSON output\nytdlp-agent download --no-proxy # bypass proxy\n```\n\n### transcript — captions as plain text\n\nDownloads VTT subtitles, strips timestamps and HTML, deduplicates lines.\n\n```bash\nytdlp-agent transcript \nytdlp-agent transcript --lang es\nytdlp-agent transcript -o D:/Transcripts --json\n```\n\n### info — metadata\n\nReturns: id, title, uploader, channel, duration, view_count, like_count, upload_date, description (first 500 chars), thumbnail, formats.\n\n```bash\nytdlp-agent info \nytdlp-agent info --json\n```\n\n### audio — MP3 extract\n\n```bash\nytdlp-agent audio \nytdlp-agent audio -o D:/Audio --json\n```\n\n### thumbnail — grab thumbnail\n\n```bash\nytdlp-agent thumbnail \nytdlp-agent thumbnail -o D:/Thumbs\n```\n\n### playlist — download playlist\n\n```bash\nytdlp-agent playlist \nytdlp-agent playlist --format mp3 -o D:/Downloads/playlist\nytdlp-agent playlist --info-only --json # metadata only\n```\n\n### channel — channel videos\n\n```bash\nytdlp-agent channel --limit 10\nytdlp-agent channel --info-only --json\n```\n\n### search — search YouTube\n\n```bash\nytdlp-agent search \"how to fix roof leak\" --limit 5 --json\n```\n\n## JSON Output Schema\n\n```json\n{\n \"ok\": true,\n \"url\": \"https://youtube.com/...\",\n \"file\": \"/path/to/output.mp4\",\n \"title\": \"Video Title\",\n \"duration_seconds\": 127,\n \"format\": \"mp4\"\n}\n```\n\n## Agent Pipeline Pattern\n\n```python\nimport subprocess, json\n\nresult = subprocess.run(\n [\"ytdlp-agent\", \"transcript\", video_url, \"--json\"],\n capture_output=True, text=True\n)\ndata = json.loads(result.stdout)\nif data[\"ok\"]:\n print(data[\"transcript\"])\n```\n\n## Related Topics\n\n- [[youtube-fetch]] — lightweight YouTube metadata + transcript sub-skill\n- [[yt-transcript-pipeline]] — automated daily transcript vault\n- [[social-transcript-extractor]] — social platform transcripts via API\n- [[assemblyai-transcription]] — AI-powered audio intelligence layer\n\n#video-sop #ytdlp #youtube #download #transcript #cli #proxy\n","html":"

yt-dlp Agent — YouTube & Video CLI

\n

Agent-native CLI wrapper for yt-dlp. Oxylabs ISP proxy by default. All commands return JSON for pipeline use.

\n

Install

\n
cd D:\\ClaudeDev\\00_GITHUB\\cliit\\clis\\ytdlp-agent\npip install -e .\n
\n

Proxy Rule

\n

All YouTube and video downloads must use the Oxylabs proxy (home IP rule). The --no-proxy flag is available for non-scraping one-off checks.

\n

Commands

\n

download — video or audio

\n
ytdlp-agent download <url>                    # MP4, current dir\nytdlp-agent download <url> -o D:/Downloads   # custom dir\nytdlp-agent download <url> --format mp3      # extract MP3\nytdlp-agent download <url> --format wav      # extract WAV\nytdlp-agent download <url> --json            # JSON output\nytdlp-agent download <url> --no-proxy        # bypass proxy\n
\n

transcript — captions as plain text

\n

Downloads VTT subtitles, strips timestamps and HTML, deduplicates lines.

\n
ytdlp-agent transcript <url>\nytdlp-agent transcript <url> --lang es\nytdlp-agent transcript <url> -o D:/Transcripts --json\n
\n

info — metadata

\n

Returns: id, title, uploader, channel, duration, view_count, like_count, upload_date, description (first 500 chars), thumbnail, formats.

\n
ytdlp-agent info <url>\nytdlp-agent info <url> --json\n
\n

audio — MP3 extract

\n
ytdlp-agent audio <url>\nytdlp-agent audio <url> -o D:/Audio --json\n
\n

thumbnail — grab thumbnail

\n
ytdlp-agent thumbnail <url>\nytdlp-agent thumbnail <url> -o D:/Thumbs\n
\n

playlist — download playlist

\n
ytdlp-agent playlist <playlist_url>\nytdlp-agent playlist <playlist_url> --format mp3 -o D:/Downloads/playlist\nytdlp-agent playlist <playlist_url> --info-only --json   # metadata only\n
\n

channel — channel videos

\n
ytdlp-agent channel <channel_url> --limit 10\nytdlp-agent channel <channel_url> --info-only --json\n
\n

search — search YouTube

\n
ytdlp-agent search \"how to fix roof leak\" --limit 5 --json\n
\n

JSON Output Schema

\n
{\n  \"ok\": true,\n  \"url\": \"https://youtube.com/...\",\n  \"file\": \"/path/to/output.mp4\",\n  \"title\": \"Video Title\",\n  \"duration_seconds\": 127,\n  \"format\": \"mp4\"\n}\n
\n

Agent Pipeline Pattern

\n
import subprocess, json\n\nresult = subprocess.run(\n    [\"ytdlp-agent\", \"transcript\", video_url, \"--json\"],\n    capture_output=True, text=True\n)\ndata = json.loads(result.stdout)\nif data[\"ok\"]:\n    print(data[\"transcript\"])\n
\n

Related Topics

\n
    \n
  • [[youtube-fetch]] — lightweight YouTube metadata + transcript sub-skill
    • \n
  • [[yt-transcript-pipeline]] — automated daily transcript vault
    • \n
  • [[social-transcript-extractor]] — social platform transcripts via API
    • \n
  • [[assemblyai-transcription]] — AI-powered audio intelligence layer
    • \n
\n

#video-sop #ytdlp #youtube #download #transcript #cli #proxy

\n"}