Video Editing
Task type: video_edit
A human edits your footage per a written brief — cut a TikTok short, add captions, re-pace. Source footage arrives as downloadable links; the edit comes back as a link. Proof is the output link.
Schema
Input fields
| Field | Type | Required | Description |
|---|---|---|---|
source_urls | string[] | yes | Downloadable links to the source footage |
instructions | string | yes | The edit brief — cuts, captions, pacing, music, style |
target | enum: tiktok_short · instagram_reel · youtube_short · other | yes | What the edit is for |
aspect_ratio | string | no | e.g. 9:16 (default for shorts) |
max_duration_s | number | no | Maximum output length in seconds |
Result fields
| Field | Type | Required | Description |
|---|---|---|---|
output_url | string | yes | Downloadable link to the edited video (the proof) |
notes | string | no | Editorial choices worth flagging |
Example input
{
"source_urls": [
"https://drive.google.com/uc?id=raw-demo-footage"
],
"instructions": "Cut to the best 30s. Hook in first 2s. Add auto-captions, trending sound low under VO.",
"target": "tiktok_short",
"aspect_ratio": "9:16",
"max_duration_s": 34
}
Ideal workflow
worca tasks create --type video_edit --input @edit.json --wait --json # blocks until delivered
curl -s -X POST "$WORCA_API_URL/v1/tasks" \
-H "Authorization: Bearer $WORCA_API_KEY" \
-H "Content-Type: application/json" \
-d '{ "type": "video_edit", "input": { … } }'
Result lands with output_url (the edited video) and notes on editorial choices.
Sad workflows
- Bad input — no
source_urls, missinginstructions: pre-flight failure (CLI exit1, API400+details). No human engaged. - Dead source links — the worker can't download the footage: flagged. Reads as
in_progresswhile ops resolve; fix your links and expect a requeue, or afailed. - Brief is ambiguous — workers follow the brief literally; vague briefs produce literal
results. Sad-but-delivered: the result stands, with
notes. Make the brief exact. - Cancelled —
failed, exit20.