ghl-video-drip - SKILL.md Agent Skill

name: ghl-video-drip version: 1.0.0 description: | Schedule N already-uploaded GHL videos as a one-per-day, multi-platform drip via a GoHighLevel Social Planner CSV. You give the CDN MP4 links (videos already in GHL Media Library) + a start date + time + captions; this builds a validated 39-column / 2-header-row Social Planner CSV that posts each clip to TikTok + Instagram Reel + YouTube Short + LinkedIn (+ a pinned lead-magnet comment), one video per day.

Distilled from batch7 (2026-05-30): 10 reels → 06-01→06-10 @ 7PM ET.

Hard-won rules baked in: - Video URL MUST sit in videoUrls (col 6), NOT gifUrl (col 5) — the col-misalignment trap that makes GHL render a clip as a broken GIF. - GHL location timezone must be America/New_York (auto-DST), NOT America/Cancun (fixed GMT-05, 1h off in summer). CSV clock-time = location TZ. - 39 columns, 2 header rows (row1 platform groups, row2 field names). - Validate with .NET TextFieldParser (Import-Csv FAILS — duplicate "type" header). - Always test row 1 in GHL before trusting the full drip.

Trigger when user says: "ghl video drip", "/ghl-video-drip", "schedule these videos one per day", "new GHL drip like batch7", or pastes N GHL CDN video links and a start date asking to schedule them.

GHL Video Drip

Turn N pre-uploaded GHL videos into a one-per-day multi-platform Social Planner CSV.

Prereqs (confirm with user)

Videos already uploaded to GHL Media Library → user has the CDN links (https://assets.cdn.filesafe.space/<loc>/media/<id>.mp4).
GHL location timezone = America/New_York (EDT) so CSV times = true ET. Set at: sub-account → Settings → Business Profile → Timezone. ⚠️ NOT America/Cancun (no DST → 1h off June–Nov). Changing TZ is global for the sub-account — warn it shifts any other queued posts/calendars; check the queue first.

Inputs to collect (use AskUserQuestion if not given)

CDN links (ordered) + a short caption/title per video.
Start date (YYYY-MM-DD), daily time (default 19:00 = 7PM ET, best US window), platforms (default all 4: TikTok + IG Reel + YT Short + LinkedIn).
Optional per-video: tags, category, pinned followUpComment (lead-magnet PDF url).

Build

Write an input JSON (see reference/sample_input.json).
Run the builder:
```
# python = your Python 3.11 interpreter
python ~/.claude/skills/ghl-video-drip/scripts/build_drip_csv.py <input.json> <out.csv>
```
- Increments date +1 day per video from start_date at time.
- Emits 39-col / 2-header CSV, all fields quoted, video in col 6.

Validate (asserts every record = 39 fields + prints date→video→title map):

powershell -File ~/.claude/skills/ghl-video-drip/scripts/validate_csv.ps1 <out.csv>

Hand the CSV path to the user → upload to GHL Social Planner → test row 1 first (confirm the video plays on all platforms, not a GIF/broken) → let the rest run.

Notes

One CSV row = one video posted to all selected platforms simultaneously (All-Social covers FB/IG/LinkedIn as reel; dedicated YouTube cols = short; TikTok cols set).
Platform selection is also gated by which social accounts are connected/picked at upload time in Social Planner.
Brand rules still apply: @yourhandle handle, client SAVINGS only (never your fees), no fake metrics.

Column map (39)

1 postAtSpecificTime · 2 content · 3 OGmetaUrl · 4 imageUrls · 5 gifUrl · 6 videoUrls · 7 mediaOptimization(TRUE) · 8 applyWatermark(FALSE) · 9 tags · 10 category · 11 followUpComment · 12-13 type=reel,reel(FB,IG) · 14-15 pdfTitle,postAsPdf · 16-25 Google GBP (empty ×10) · 26 YT title · 27 YT privacy(public) · 28 YT type(short) · 29 TikTok privacy(everyone) · 30 promoteOtherBrand(FALSE) · 31-33 enableComment/Duet/Stitch(TRUE) · 34 videoDisclosure(FALSE) · 35 promoteYourBrand(FALSE) · 36-37 Community title,notify · 38-39 Pinterest title,link.