Overview
A comprehensive media processing service that transforms images and videos through both simple UI-driven operations and advanced transform scripts. The service uses a JSON DSL (Domain Specific Language) - a structured scripting language that allows users to define multi-step media transformations as a sequence of operations, and is designed to be easily generated by LLMs. These transform scripts can be hand-written by users or dynamically generated by LLMs, making the service ideal for AI-orchestrated media workflows with support for batch processing, scene detection, and intelligent transformations.
This service is in its initial, experimental form, with a limited set of functions and transform scripting language cabilities as to allow for meaningful experiments
into the use of LLMs for building AI-assisted media workflows, in film and video and other media production settings. If it proves worthwhile, we will significantly extend the
set of functions it can perform and the controls provided.
Service Information
| Property |
Value |
| Service Name |
HyperFlow Media Transform |
| Status |
Enabled |
| Compatible Nodes |
Transform Content |
Key Features
- Dual operation modes: Simple UI for single operations, Advanced mode for complex transform scripts
- Multi-input support: Process up to three media sources simultaneously
- Internal media workspace: Allows the creation and storing of any number of intermediate media items in a scripted workflow
- Batch processing: Automatically handles Import Sets with multiple items
- Scene intelligence: AI-ready scene detection and frame extraction
- LLM integration: Transform scripts designed for easy generation by AI systems
- Key-value store: Can store media references persistently for use across different flow-graphs or multiple flow-graph runs.
Inputs
| Input |
Type |
Required |
Description |
| Primary media |
Content/Import Set |
Yes |
Main media file(s) to transform |
| Secondary media |
Content/Import Set |
No |
For compositing, watermarking operations |
| Additional media |
Content/Import Set |
No |
Third media input for complex workflows |
Parameters
Mode Selection
| Parameter |
Type |
Default |
Description |
| Mode |
Choice |
Simple |
"Simple" for UI operations, "Advanced" for transform scripts |
Simple Mode Operations
Image Operations