Step 3 — Convert Idea to Vidsheet
You have ideas from Step 2. Each idea has a video_type (talking avatar, montage, split-screen, etc.) and a project_manifest describing the timeline. Now you turn that idea into an Auto Content Engine (vidsheet) — the actual production surface where rows, columns, cells, and layers live.
There are two paths. Both are legitimate, but one is dramatically faster.
Path A — Clone a template (recommended)
Section titled “Path A — Clone a template (recommended)”Templates are pre-configured engines with the right columns, creation cards, and layer stacks for a specific video type. Cloning a template gives you a working engine in one API call.
curl -X POST "https://api.gen.pro/v1/templates/spreadsheets/tiktok-talking-avatar/clone" \ -H "X-API-Key: $GEN_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "agent_id": "'"$AGENT_ID"'" }'Returns a fully configured engine with columns for script, voice, video, and final output — ready to fill. Jump to Start from a Template for the full flow.
Path B — Build a custom engine
Section titled “Path B — Build a custom engine”If no template matches your idea, build an engine from scratch:
curl -X POST "https://api.gen.pro/v1/autocontentengine?agent_id=$AGENT_ID" \ -H "X-API-Key: $GEN_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "spreadsheet": { "title": "Custom campaign" } }'Then create each column, configure creation cards, and set up layers by hand. See Build a Custom Engine. Expect 15–25 API calls vs the 1 for a template clone.
Match the idea to a template
Section titled “Match the idea to a template”Each idea has a video_type_id (see video type IDs). List templates and pick one whose tags include the same video type:
curl "https://api.gen.pro/v1/templates/projects" \ -H "X-API-Key: $GEN_API_KEY"Response includes slug, title, description, tags, and cover. Pick a slug that matches the idea’s video type, then clone.
Video type IDs
Section titled “Video type IDs”| ID | Name | Typical templates |
|---|---|---|
| 0 | Any | Generic engines |
| 1 | Talking Avatar | AI talking head + lipsync + captions |
| 2 | Green Screen | Presenter over dynamic background |
| 3 | Montage | Short clips stitched with music |
| 4 | Text-Driven | Kinetic text + voiceover |
| 5 | POV Object | Object-focused (unboxing, product shots) |
| 6 | Narrated / Voiceover | VO + b-roll |
| 8 | Split Screen | Reaction / comparison format |
| 9 | Skit | Multi-character scene |