Global Tool Fields (at the top of api.yaml)

Field	Description
name	Unique identifier for the tool
description	Detailed explanation of the parameter's purpose and usage
tip	Additional advice or context for using the parameter effectively (targetted to help AI agents use the tool well)
output_type	Output type of the tool (image, video, audio)
handler	What backend the tool is running on
comfyui_output_node_id	ComfyUI output node id if applicable
private	Optional boolean indicating whether to hide the tool from the Eden app UI

Parameter Fields

Field	Description
name	Unique identifier for the parameter
label	Human-readable name for the parameter
description	Detailed explanation of the parameter's purpose and usage
tip	Additional advice or context for using the parameter effectively (targetted to help AI agents use the tool well)
type	Data type of the parameter (e.g., string, int, float, bool, image, video, image[], lora, audio)
default	Default value for the parameter if not specified
minimum	Minimum allowed value for numeric types
maximum	Maximum allowed value for numeric types
min_length	Min length of array types
max_length	Max length of array types
step	For numeric types, specifies the increment step in the UI (e.g., 8 for width/height)
choices	List of predefined options for the parameter, if applicable
Special Fields
hide_from_agent	Boolean indicating whether to hide the parameter from AI agents (true/false), defaults to false when not provided
hide_from_ui	Boolean indicating whether to hide the parameter from the Eden webapp UI (true/false), defaults to false when not provided
required	Boolean indicating if the parameter is mandatory (true/false), defaults to false when not provided
comfyui:
node_id	Identifier for the ComfyUI node
field	Field within the node
subfield	Specific subfield within the node field
preprocessing	Any preprocessing steps required to properly parse this input (e.g., "folder" for image arrays, which saves a set of imgs to a tmp folder and passes that path in ComfyUI)

Remapping labels to multiple nodes example:

- name: preprocessor1
  label: Controlnet1 preprocessor
  description: Type of controlnet preprocessor. Examples can be seen at https://github.com/lllyasviel/ControlNet-v1-1-nightly?tab=readme-ov-file#controlnet-11-depth
  tip:  depth will try to maintain the perceived depth of the input scene. Canny edge creates strong edges adhering to the shape of your image, whereas scribble will create guidance towards a rougher sketched shape of your starting image and often produces better quality video at the cost of less finegrained correspondence with the control image. Pose will try to extract the pose from a person and inject it into the video.
  type: string
  visible_if: use_controlnet1=true
  default: CannyEdgePreprocessor
  choices: [CannyEdgePreprocessor, DepthAnythingV2Preprocessor, AnyLineArtPreprocessor_aux, DensePosePreprocessor, Scribble_XDoG_Preprocessor, none]
  choices_label: [Edges (Canny), Depth, Lineart, human pose, Scribble lines, Luminance (QR dark/bright patterns)]
  comfyui: 
    node_id: 406
    field: inputs
    subfield: preprocessor    
    remap:
      - node_id: 107
        field: inputs
        subfield: control_net_name
        value:
          - input: CannyEdgePreprocessor
            output: control_v11p_sd15_canny.pth
          - input: DepthAnythingV2Preprocessor
            output: control_v11f1p_sd15_depth.pth
          - input: AnyLineArtPreprocessor_aux
            output: control_v11p_sd15_lineart.pth
          - input: DensePosePreprocessor
            output: control_v11p_sd15_openpose.pth
          - input: Scribble_XDoG_Preprocessor
            output: control_v11p_sd15_scribble.pth
          - input: none
            output: controlnetQRPatternQR_v2Sd15.safetensors