onboard

Onboarding — Content Repurpose System

This is Step 0 of the system. Run it once. It personalizes every other skill to one creator so the titles, hooks, blog posts, and clips come out in their voice and grounded in their channel's data — not generic AI slop.

Be conversational and fast. Ask for what you need, write the config, confirm, and hand off to the workflow. Do not lecture.

What this sets up

By the end the creator will have:

1. Their channel connected — so yt-titles and content-cascade can read their real upload history and "latest video."
2. Their Voice DNA written into the prompts — so content-cascade writes blog/Twitter/LinkedIn in their voice, not a template voice.
3. Optional integrations — Apify (reliable transcripts), Supabase (auto-save drafts), Twitter (auto-post threads). All skippable.

Everything is stored inside the plugin at ${CLAUDE_PLUGIN_ROOT}, so it survives across sessions.

Step 1 — Connect the channel (required)

Ask: "What's your YouTube channel handle or URL?" (e.g. @SaminYasar_ or https://www.youtube.com/@SaminYasar_)

Normalize whatever they give you to a handle (@SaminYasar_) and a videos URL (https://www.youtube.com/@SaminYasar_/videos).

Then write it into the two places the skills read from:

a) The performance script — replace the CHANNEL_URL constant:

python3 - "$CLAUDE_PLUGIN_ROOT" "@THEIR_HANDLE" <<'PY'
import re, sys, pathlib
root, handle = sys.argv[1], sys.argv[2]
p = pathlib.Path(root, "skills/yt-titles/scripts/channel_performance.py")
url = f"https://www.youtube.com/{handle}/videos"
text = p.read_text()
text = re.sub(r'CHANNEL_URL = "[^"]*"', f'CHANNEL_URL = "{url}"', text)
p.write_text(text)
print("Channel set to", url)
PY

b) The channel ID used for "latest video" auto-detect. The content-cascade and short-form skills auto-detect the latest upload with yt-dlp. A handle works directly — resolve and store the canonical channel ID once so it's unambiguous: