ML Pipeline Workflow
Guides an agent in designing end-to-end MLOps pipelines from data prep through training, validation and deployment
- Repository
- wshobson/agents
- Stars
- 42,095
- Licence
- MIT
- Files
- 1 file
What it does
This skill gives the agent a structured playbook for building machine learning pipelines that take data through preparation, model training, validation, and production deployment with monitoring. Once installed, the agent can suggest pipeline stages, orchestration patterns (such as Airflow, Dagster, or Kubeflow), and deployment strategies like canary or blue-green releases, and it can point to bundled templates and reference guides. The file itself is mostly high-level guidance and lists of tools; it does not contain executable code.
Great for
- Engineers building their first production ML pipeline who want a stage-by-stage outline
- Data teams standardising how models move from training to deployment
- Developers who want to compare orchestration and serving options before choosing one
Skip it if
- Non-technical users with no ML infrastructure: the advice assumes existing cloud, orchestration and model-serving setups, and the skill does not run anything for you
- Teams needing a verified, tested implementation: the bundled references and templates are cited but not included in the provided file, so their content could not be checked
Try asking
- “Design a batch training pipeline for a churn prediction model with validation and rollback steps”
- “Compare Airflow, Dagster and Kubeflow for orchestrating our nightly model retraining”
- “Outline a canary deployment plan for a new recommendation model on Kubernetes with KServe”
Install in 3 steps
- 1
Download the skill
The ZIP contains the 1 original files, the MIT licence and an attribution note.
ml-pipeline-workflow.zip - 2
Add it to your agent
Claude Code (all projects)
mkdir -p ~/.claude/skills && unzip ml-pipeline-workflow.zip -d ~/.claude/skills/Claude Code / Agent SDK (this project only)
mkdir -p .claude/skills && unzip ml-pipeline-workflow.zip -d .claude/skills/Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.
- 3
Ask for the task
No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.
Get better results
- Name your cloud platform and orchestration tool up front so the advice matches your stack
- Ask for a staged rollout plan with explicit validation gates rather than a generic overview
- Check any generated configuration against your own infrastructure, since the skill gives patterns rather than tested code
What's inside
- SKILL.md
Redistributed unmodified from wshobson/agents under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.