Quick overview
This self-hosted workflow exposes a webhook that takes a product image path or URL, uses the insMind CLI to select an image-to-video tool and submit an async character video generation job, then returns the resulting video URL (or a pending/failed status).
How it works
- Receives a POST request via webhook containing a product image path or URL (and optional video settings).
- Checks that insMind CLI is authenticated on the n8n host and queries insMind CLI for the available tools and models.
- Builds a prompt and selects a compatible insMind CLI video tool (or uses the provided insmind_tool) and insmind-cli call command for async generation.
- Retrieves the selected tool details and submits the async generation job via insmind-cli.
- Extracts the returned task ID, waits briefly, then fetches the task status/result from insmind-cli.
- Parses the task output to find a video result URL and responds to the webhook with JSON including success status, workflow status, and the result URL when available.
Setup
- Use a self-hosted n8n instance that allows Execute Command, and install insmind-cli on the same worker/container where the workflow runs.
- Authenticate insmind-cli on that machine/container user (for example, run insmind-cli auth login) and verify insmind-cli auth status succeeds.
- Activate the workflow and copy the webhook URL, then POST a body that includes product_image_path or product_image_url (optionally insmind_tool, duration_seconds, aspect_ratio, product_name, campaign_goal, and story_language).
Requirements
Customization
- You can customize product_name, campaign_goal, story_language,
- duration_seconds, aspect_ratio, and insmind_tool in the webhook body. If
- insmind_tool is not provided, the workflow discovers available insMind CLI
- tools at runtime and chooses a compatible video generation tool.
Additional info
insMind CLI is useful when you want AI image or video generation inside
automated workflows without manually operating a web editor. This template
uses the CLI for tool discovery, async task submission, and task result
retrieval, making it suitable for product-image-to-video automation in self-
hosted n8n.