amc-run-sample-calibration - Calibrate AMC with the bundled sample dataset
Runs end-to-end calibration on the bundled sample dataset against a running AMC microservice and reports evaluation metrics.
Tags
Updated: 2026-09-28Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Detect running AMC backend
- Extract sample calibration data
- Create calibration project
- Upload sample files
- Run AMC calibration
- Poll project status
- Report evaluation metrics
- Attempt VGGT refinement
Inputs
- AMC microservice
- AutoMagicCalib checkout
- Sample dataset ZIP
- Backend port
- Python environment
Outputs
- Calibration project
- Calibration status
- Evaluation metrics
- VGGT refinement status
- Manual inspection UI URL
Requirements
- AMC microservice running on port 8000-8009
- Sample ZIP available at assets/sdg_08_2_sample_data_010926.zip
- Python 3 with requests or Swagger UI access
- Access to the AutoMagicCalib checkout