MateAI YouTube Studio & Strategy Suite (yt-studio)

Built by MateAI (@mateai-c6m) · GitHub Repo.

Shoutout & Credits: This plugin combines MateAI's local YouTube Studio & Analytics MCP Server with the open-source YouTube strategy & creator workflows by Jeremy Longshore (@jeremylongshore / youtube-strategy) and Daniel Agrici (@AgriciDaniel / claude-youtube).

End-to-End AI YouTube Workflow (Strategy Skills + Live MCP Execution)

Whenever planning, scripting, packaging, uploading, or analyzing a YouTube video or Short, combine the bundled strategy skills with the live mateai-youtube MCP tools:

When to Use This Skill

Use this skill whenever the user asks to:

• Check YouTube channel views, subscriber growth, or video performance (for @mateai-c6m or any public @handle).
• Analyze YouTube Studio retention (Avg % Viewed), Shorts hook performance (Viewed vs. Swiped Away %), Engaged Views, or Subscribers Gained.
• Update the Shorts A/B/C visual style experiment table in pipeline/explainer-short/README.md.
• Upload a new video or Short (MP4/MOV), attach a custom cover (PNG/JPG), schedule a publish date/time (publish_at in ISO 8601), or post a pinned comment.
• Edit an existing video's title, description, tags, privacy status, or thumbnail.

Available MCP Tools (mateai-youtube)

1. yt_channel_snapshot(handle="@mateai-c6m", force_public=false): Pulls live Studio OAuth2 metrics (without OAuth it returns setup instructions; pass force_public=true only if the user asks for public-page scraping) and writes report.html.
2. yt_shorts_experiment_table(handle="@mateai-c6m"): Returns a formatted Markdown table comparing Shorts performance.
3. yt_video_details(video_id): Returns metadata and stats for a single 11-char video ID.
4. yt_upload_draft(video_path, spec_path, thumbnail_path, privacy_status="private", publish_at=None, auto_post_comment=false): Uploads a rendered video + Markdown spec (## Title, ## Description, ## Tags, ## Pinned comment) to YouTube Studio as a draft or scheduled upload.
5. yt_update_video(video_id, title, description, tags, privacy_status, publish_at, thumbnail_path, contains_synthetic_media): Updates packaging or schedule of an existing video. When changing privacy_status or publish_at, also pass contains_synthetic_media (true or false); ask the user if you don't know it.
6. yt_post_comment(video_id, text): Posts a top-level comment on a video.

TikTok tools (same server)

7. tiktok_channel_snapshot(): Official TikTok Display API stats for the connected account (public videos only).
8. tiktok_upload_draft(video_path, spec_path, title, hashtags): Sends a Short to the TikTok inbox as a draft; the owner finishes the post in the app.
9. tiktok_draft_status(publish_id): Read-only. Where an earlier TikTok upload stands (inbox, processing, posted, or failed with TikTok's reason).
10. cross_platform_shorts_table(yt_handle, yt_force_public): YouTube Shorts vs TikTok comparison table.

For rules, limits and what is not built yet, read skills/tiktok-studio/SKILL.md.

CLI Fallback Commands

If running via terminal from /Users/mategog/Documents/Yt/MateAI/builds/03-yt-studio-mcp:

# 1. Fetch live Studio OAuth snapshot & render report.html
python3 yt_studio.py snapshot

# 2. Generate Shorts Style Experiment Markdown Table
python3 yt_studio.py experiment-table

# 3. Re-authenticate or switch YouTube Brand Channel
python3 yt_studio.py auth

# 4. Upload a rendered Short/Video + upload spec as a Private Draft in Studio
python3 yt_studio.py upload-draft \
  --video ../../exports/gemini-4-carbon-poster.mp4 \
  --spec ../../planning/gemini-4-carbon-upload.md \
  --thumbnail ../../assets/thumbnails/gemini-4-carbon-poster.png