ml-pipeline-automation

Automate ML workflows with Airflow, Kubeflow, MLflow. Use for reproducible pipelines, retraining schedules, MLOps, or encountering task failures, dependency errors, experiment tracking issues.

1.25x

Quality

82%

Does it follow best practices?

Impact

100%

1.25x

Average score across 3 eval scenarios

Securityby

Low

Low-risk findings worth noting

Quality

Content

65%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is highly actionable with real, executable examples and well-structured references, but it is verbose with redundant code blocks and time-sensitive version pinning, and the main workflow lacks explicit validation checkpoints. Tightening inline content and pushing detail into references would lift the weaker dimensions.

Suggestions

Collapse the duplicate DAG examples (Quick Start and 'Basic Airflow DAG') into a single canonical example to remove redundancy and reduce tokens.

Move time-sensitive version pinning into a dedicated 'versions' note or reference file, or add a 'verify latest' step instead of hardcoding dated versions in the main body.

Push the seven 'Known Issues Prevention' entries into a reference file (e.g. references/airflow-patterns.md) and keep only a brief pointer in SKILL.md to improve progressive disclosure.

Add explicit validation/verification checkpoints (e.g., verify DAG parses, confirm MLflow run logged) to the Quick Start workflow to strengthen workflow clarity.

Dimension	Reasoning	Score
Conciseness	The ~420-line body is padded with two near-duplicate DAG examples and seven verbose problem/solution entries, and pins time-sensitive versions ('apache-airflow==3.1.5 mlflow==3.7.0', 'current as of December 2025') which penalize conciseness; it could be tightened significantly.	2 / 3
Actionability	Provides concrete, executable code and shell commands throughout (Quick Start DAG, validation tasks, alerting callbacks, sensors, TaskGroups) that are copy-paste ready.	3 / 3
Workflow Clarity	The Quick Start lists five sequenced steps and validation appears in data-quality tasks, but the main training workflow lacks explicit validation checkpoints and the Basic DAG uses a placeholder ('# model.fit(X, y) ...') rather than a complete sequence.	2 / 3
Progressive Disclosure	References are clearly signaled one level deep with explicit 'Load ... when' guidance and real reference files exist, but the SKILL.md itself is monolithic with large inline sections (e.g., all seven 'Known Issues' entries) that would be better split into the reference files.	2 / 3
	Total	9 / 12 Passed

Description

100%

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is specific, tool-named, and includes an explicit 'Use for ...' trigger clause covering both what the skill does and when to use it. It is concise yet comprehensive and uses third-person voice throughout.

Dimension	Reasoning	Score
Specificity	Names the concrete tools (Airflow, Kubeflow, MLflow) and lists multiple specific actions such as 'reproducible pipelines, retraining schedules, MLOps' rather than vague language.	3 / 3
Completeness	Explicitly answers both 'what' ('Automate ML workflows with Airflow, Kubeflow, MLflow') and 'when' via a clear 'Use for ...' trigger clause, matching the top anchor.	3 / 3
Trigger Term Quality	Includes natural phrases a user would say — 'reproducible pipelines, retraining schedules, MLOps, task failures, dependency errors, experiment tracking issues' — covering common variations.	3 / 3
Distinctiveness Conflict Risk	Targets a clear niche (ML pipeline orchestration with named tools) and is unlikely to trigger for unrelated skills.	3 / 3
	Total	12 / 12 Passed

Validation

93%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
metadata_version	'metadata.version' is missing	Warning

	Total	15 / 16 Passed

Repository: secondsky/claude-skills
Path: plugins/ml-pipeline-automation/skills/ml-pipeline-automation/SKILL.md
Commit: c4889f6

Reviewed: about 2 hours ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.