Model Deployment Helper Skill for Local Prediction Contracts
Serve your trained artifact in a local endpoint and compare its input validation and prediction with the offline model.
中文Install in Claude Code
Use your own model artifact, Python, and the project's web framework. The fixture predicts packing duration.
- Get the complete `plugins/ai-ml/model-deployment-helper/skills/deploying-machine-learning-models` directory from the pinned GitHub commit.
- Copy it into the target project's `.claude/skills/`. Check `.claude/skills/deploying-machine-learning-models/SKILL.md`, start Claude Code at the project root, and call `/deploying-machine-learning-models`.
Example task
POST item_count=2 and parcel_size=4 to local /predict. Repeat without parcel_size. Compare the valid result with the same model's offline prediction.What to check
The valid response has the expected value and model version. The invalid request returns a field validation error.
Original SKILL.md on GitHub · Updated
Frequently asked question
Can the service load an arbitrary model file?
Use a pinned artifact you generated or trust, following the format's safe loading guidance.