Data labeling and annotation platform for ML training data.
## Label Studio MCP Server: Open Source Data Labeling The **Label Studio MCP Server** integrates Label Studio's data annotation platform into Google Antigravity. This open-source solution supports labeling for images, text, audio, video, and time series, making it essential for building high-quality ML training datasets. ### Why Label Studio MCP? Label Studio simplifies the data labeling workflow: - **Multi-Modal**: Label images, text, audio, video, and more - **Customizable**: Build any labeling interface with templates - **ML Integration**: Pre-labeling with model predictions - **Collaborative**: Team-based annotation workflows - **Open Source**: Self-hosted with enterprise features ### Key Features #### 1. Flexible Labeling Interfaces ```xml <!-- Custom labeling interface for NER --> <View> <Labels name="ner" toName="text"> <Label value="Person" background="red"/> <Label value="Organization" background="blue"/> <Label value="Location" background="green"/> </Labels> <Text name="text" value="$text"/> </View> <!-- Image classification with choices --> <View> <Image name="image" value="$image"/> <Choices name="choice" toName="image"> <Choice value="Cat"/> <Choice value="Dog"/> <Choice value="Other"/> </Choices> </View> ``` #### 2. ML Backend Integration ```python from label_studio_ml import LabelStudioMLBase class MyModel(LabelStudioMLBase): def predict(self, tasks, **kwargs): predictions = [] for task in tasks: # Use your model for pre-labeling result = self.model.predict(task["data"]["text"]) predictions.append({ "result": [{ "from_name": "ner", "to_name": "text", "type": "labels", "value": {"labels": [result]} }] }) return predictions ``` #### 3. API Access ```python from label_studio_sdk import Client ls = Client(url="http://localhost:8080", api_key="your-key") # Create project project = ls.start_project( title="Text Classification", label_config=label_config ) # Import tasks project.import_tasks([ {"text": "This is great!"}, {"text": "Not happy with this."} ]) # Export annotations annotations = project.export_tasks() ``` ### Configuration ```json { "mcpServers": { "label-studio": { "command": "npx", "args": ["-y", "@anthropic/mcp-label-studio"], "env": { "LABEL_STUDIO_URL": "http://localhost:8080", "LABEL_STUDIO_API_KEY": "your-api-key" } } } } ``` ### Use Cases **Training Data Creation**: Build high-quality labeled datasets for supervised machine learning models. **Active Learning**: Use model predictions to prioritize uncertain samples for human review. **Quality Assurance**: Multi-annotator workflows with consensus and quality metrics. The Label Studio MCP Server enables efficient data labeling workflows within Antigravity.
{
"mcpServers": {
"label-studio": {}
}
}