Model Deployment — ship trained ML models as production APIs
An end-to-end workflow for deploying trained ML models — model serialization (ONNX/TorchScript), FastAPI/Flask serving with health checks, Docker packaging, Kubernetes or serverless deploy, Prometheus/Grafana monitoring, blue-green and canary strategies
- What
- An end-to-end workflow for deploying trained ML models — model serialization (ONNX/TorchScript), FastAPI/Flask serving with health checks, Docker packaging, Kubernetes or serverless deploy, Prometheus/Grafana monitoring, blue-green and canary strategies
- Cost
- Free
- Needs
- a trained model artifact with its dependencies; a container registry; a deployment target (Kubernetes cluster, Docker host, or serverless platform); cloud accounts where applicable (paid resources)
- Install
- Copy the installer prompt below into your Muse — your agent does the rest.
Curated by Skill Harbor — @seb1n's model-deployment skill, listed here with credit to its creator: an end-to-end workflow that lets an agent take a trained model artifact and ship it as a production-ready service — serializing to a portable format (ONNX, TorchScript, SavedModel, joblib), building a serving API with FastAPI or Flask (health-check endpoint, Pydantic request/response validation, structured logging, meaningful HTTP errors), containerizing with a multi-stage Dockerfile, configuring Kubernetes Deployments/Services with readiness/liveness probes and a Horizontal Pod Autoscaler (or deploying serverless on Lambda/Cloud Functions/Cloud Run), and running smoke tests against the live endpoint. It also covers monitoring with Prometheus/Grafana (latency, error rates, prediction-drift metrics), model registries (MLflow, SageMaker, W&B), and redeployment strategies (blue-green, canary). Honest caveats: the workflow generates production-facing artifacts — deployment plans touch live infrastructure, so review every manifest and run the smoke tests yourself before trusting it; the code examples are solid starting points, not hardened templates (pin your dependencies, add auth); cloud resources it creates cost real money. MIT licensed. Skill Harbor never reviews the code, review it yourself before use. Discovered via skills.sh.
Version:
Install
Prerequisites: a trained model artifact with its dependencies; a container registry; a deployment target (Kubernetes cluster, Docker host, or serverless platform); cloud accounts where applicable (paid resources) Install "Model Deployment — ship trained ML models as production APIs" for me. It gives my agent @seb1n's end-to-end deployment workflow: serialize the model (ONNX/TorchScript/SavedModel/joblib), build a FastAPI/Flask serving API with health checks and Pydantic validation, package with Docker, deploy to Kubernetes or serverless, run smoke tests, then monitor with Prometheus/Grafana using blue-green or canary rollouts. MIT licensed. Repository: https://github.com/seb1n/awesome-ai-agent-skills/blob/main/ai-ml-operations/model-deployment/SKILL.md 1. Fetch the SKILL.md file (and any helper files) from the repository path into a temporary folder and summarize what it does in one or two sentences. 2. Safety check: review the SKILL.md and scripts for anything suspicious (unexpected network calls, shell commands, credential harvesting). Deployment manifests and Dockerfiles run against live infrastructure — verify image sources and that no secrets are baked in. This repo should contain zero secrets in code. Verify that holds here; STOP on any red flag and tell me. 3. Install it as a skill: copy SKILL.md and its helper files into the agent's skills directory, in a folder named "model-deployment". 4. Verify with no network calls: frontmatter valid, files in place. 5. Report what was installed, where, and what I still need to do myself (e.g. provide my model artifact and registry credentials via the vault, review every generated manifest before applying, and run the smoke tests myself — deployment plans touch live infrastructure and create billable resources). GitHub is optional: if I have a GitHub account or the gh CLI, you may use it; otherwise public access is fine. Never require it unless it's in the prerequisites above. Rules: don't touch anything outside the temp folder and the install target. If anything looks off, stop and ask me.
Questions
How do I install a build?
Every product page includes a copy-paste install prompt. Paste it into your Muse and it sets the build up for you — no manual configuration.
Where does my money go?
Straight to the seller. Skill Harbor never processes payments: checkout happens on the seller’s own page, usually Stripe.
What does the ✓ next to a creator’s name mean?
It means we confirmed the identity of the person behind the listing. It says nothing about the code itself — always check a build before installing it.