Roboflow Asset Library Attributes

Update attributes and tags for Asset Library images.

Submit attribute and tag updates for existing Asset Library images, enabling enrichment workflows where model outputs become filterable image fields.

How This Block Works

This block submits key-value attributes and tags for existing Asset Library images in your Roboflow workspace. Attribute values are stored as image metadata. The block:

1. Receives Asset Library source image IDs, optional attributes, and optional tags

2. Resolves the target workspace from the configured Roboflow API key

3. Skips rows where both attributes and tags are empty

4. Merges duplicate source IDs using sequential semantics: later attribute values win, and tags are added as a de-duplicated set

5. Submits one batch update request and returns one submission status per input source ID, in input order

Re-running the same workflow against the same source IDs is safe: attribute keys are upserted (last write wins) and tags are unioned. There is no destructive write.

The block does not send image bytes and does not create new images. It only updates existing Asset Library source images. Removing attribute keys, removing tags, writing annotations, and creating images are intentionally out of scope for this workflow block.

Requirements

This block requires a valid Roboflow API key. The API key determines the workspace whose Asset Library images can be updated.

Type identifier

Use the following identifier in step "type" field: roboflow_core/asset_library_attributes@v1 to add the block as a step in your workflow.

Properties

Name	Type	Description	Refs
name	str	Enter a unique identifier for this step..	❌
source_id	str	Asset Library source image ID to update. For batch workflows, provide one source ID per image..	✅
metadata	Dict[str, Union[bool, float, int, str]]	Optional key-value attributes to set on the Asset Library image. Attributes are stored as image metadata. Either an inline dict whose values may be static or selector references (e.g. $inputs.camera_id), or a whole-field selector to a per-row dict produced by an upstream step..	✅
tags	List[str]	Optional tags to add to the Asset Library image. Each entry may be a static string or a reference to a workflow input/step (e.g. $inputs.label)..	✅
disable_sink	bool	If True, the block execution is disabled and no Asset Library attribute writes occur..	✅

The Refs column marks possibility to parametrise the property with dynamic values available in workflow runtime. See Bindings for more info.

Runtime compatibility

requires_internet - air-gapped / offline deployments : This block depends on a service that is not reachable from fully offline / air-gapped deployments.

Input and Output Bindings

The available connections depend on its binding kinds. Check what binding kinds Roboflow Asset Library Attributes in version v1 has.

Input and output bindings

• input

  • source_id (string): Asset Library source image ID to update. For batch workflows, provide one source ID per image..

  • metadata (Union[*, dictionary]): Optional key-value attributes to set on the Asset Library image. Attributes are stored as image metadata. Either an inline dict whose values may be static or selector references (e.g. $inputs.camera_id), or a whole-field selector to a per-row dict produced by an upstream step..

  • tags (Union[list_of_values, string]): Optional tags to add to the Asset Library image. Each entry may be a static string or a reference to a workflow input/step (e.g. $inputs.label)..

  • disable_sink (boolean): If True, the block execution is disabled and no Asset Library attribute writes occur..

• output

  • error_status (boolean): Boolean flag.

  • message (string): String value.

Example JSON definition

{
	    "name": "<your_step_name_here>",
	    "type": "roboflow_core/asset_library_attributes@v1",
	    "source_id": "$inputs.source_id",
	    "metadata": {
	        "color": "red",
	        "score": 0.98
	    },
	    "tags": [
	        "auto-labeled",
	        "red"
	    ],
	    "disable_sink": false
	}