Skills ml-engineer
📦
ml-engineer
Content revision r2 Safe ⚙️ External commands
Build Production ML Systems
Production machine learning requires reliable architecture, deployment, testing, and monitoring. This skill provides practical guidance across the complete ML system lifecycle.
Supports: Claude Codex Code(CC)
Install with my Agent
Copy this request to your Agent. It includes the canonical Skill page and manifest.
Agent request
Review the Skillstore skill "ml-engineer" from https://skillstore.io/skills/sickn33-ml-engineer.md and its manifest at https://skillstore.io/api/skills/sickn33-ml-engineer/manifest. Verify the artifact. You may proceed after verification, subject to the environment's own policy.Your Agent should still show its plan and request any confirmation required by the security policy.
Agent-readable resources
Use these links when an AI agent, crawler, or script needs clean context instead of reading the full page.
Test it
Using "ml-engineer". Design real-time serving for a fraud model with strict latency and safe rollbacks.
Expected outcome:
- Use a stateless inference service with cached online features and bounded request timeouts.
- Track latency, errors, feature freshness, score distribution, and confirmed fraud outcomes.
- Deploy with shadow traffic, then a canary rollout with automatic rollback thresholds.
Using "ml-engineer". Plan monitoring for a daily demand forecasting pipeline.
Expected outcome:
- Validate input completeness, schema, ranges, and freshness before each forecast run.
- Measure forecast error by horizon, region, and product after actual values arrive.
- Trigger investigation when data drift, repeated job failures, or sustained error thresholds occur.
Using "ml-engineer". Reduce inference cost while protecting model quality.
Expected outcome:
- Profile compute, memory, serialization, and queue time before choosing an optimization.
- Evaluate batching, quantization, caching, autoscaling, and smaller models against a fixed benchmark.
- Set quality, latency, and cost acceptance thresholds before rollout.
Security Audit
Safev5 • 8/4/2026 Open versioned report
Both static findings are false positives caused by ordinary Markdown prose. The skill contains no executable shell command, system reconnaissance behavior, or semantic evidence of malicious intent.
1
Files scanned
172
Lines analyzed
0
Review items
0
False positives ignored
Risk Factors
⚙️ External commands (1)
No confirmed security findings were detected by the latest completed static and semantic audit. This does not prove the skill has no side effects.
Audited by: codex View Audit History →
Share & cite this report
Share the versioned assessment report, neutral badge, embed card, and citations. Skillstore reports evidence without deciding whether this Skill is safe.
Copy report link
https://skillstore.io/skills/sickn33-ml-engineer/audits/5?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_reportMarkdown badge
[](https://skillstore.io/skills/sickn33-ml-engineer?utm_source=security_passport_badge)HTML badge
<a href="https://skillstore.io/skills/sickn33-ml-engineer?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/sickn33-ml-engineer/security.svg" alt="Skillstore security assessment" loading="lazy"></a>Embed card
<iframe src="https://skillstore.io/embed/skills/sickn33-ml-engineer.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>Academic citations (APA · BibTeX · CFF)
APA citation
sickn33. (2026). ml-engineer security audit report (audit version 5) [Author version unspecified]. Skillstore. https://skillstore.io/skills/sickn33-ml-engineer/audits/5BibTeX citation
@techreport{sickn33-sickn33-ml-engineer-2026,
author = {sickn33},
title = {ml-engineer security audit report (audit version 5)},
institution = {Skillstore},
year = {2026},
number = {5},
url = {https://skillstore.io/skills/sickn33-ml-engineer/audits/5},
note = {Author version unspecified}
}CITATION.cff
cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "ml-engineer security audit report (audit version 5)"
version: "unspecified"
type: report
authors:
- name: "sickn33"
date-released: "2026-08-04"
url: "https://skillstore.io/skills/sickn33-ml-engineer/audits/5"
identifiers:
- type: other
value: "skillstore:sickn33-ml-engineer:audit:5"
description: "Skillstore immutable audit report identifier"
Skillstore Score
Why this score Evidence Confidence: High 55
Architecture
85
Maintainability
87
Content
69
Community
83
Spec Compliance
What You Can Build
Plan a first production deployment
Select a serving pattern, deployment platform, monitoring plan, and rollback strategy for an existing model.
Scale an ML platform
Design distributed training, feature serving, experiment tracking, and continuous delivery for growing workloads.
Improve model operations
Define drift alerts, business metrics, retraining triggers, and controlled model rollout procedures.
Try These Prompts
Choose a deployment approach
Recommend a deployment approach for my [model type]. It serves [traffic] with [latency target] on [platform]. Explain tradeoffs and validation steps.
Design a monitored inference service
Design a [batch or real-time] inference service for [workload]. Include APIs, scaling, failure handling, metrics, alerts, testing, and rollback.
Create an experiment and rollout plan
Create an A/B test for [model change] using [business metric]. Define assignment, sample sizing, guardrails, significance checks, rollout, and stopping rules.
Architect an end-to-end ML platform
Architect an ML platform for [use case] at [scale]. Cover data, features, training, registry, serving, observability, governance, cost, and disaster recovery.
Best Practices
- Provide workload scale, latency, cost, data, platform, and compliance constraints.
- Validate recommendations with representative data, load tests, failure tests, and staged rollouts.
- Measure business outcomes with model quality, reliability, fairness, and infrastructure metrics.
Avoid
- Do not optimize model accuracy while ignoring latency, reliability, and business impact.
- Do not deploy without versioning, observability, rollback procedures, and ownership.
- Do not automate retraining without data validation, evaluation gates, and approval controls.
Frequently Asked Questions
Which ML frameworks does this skill cover?
It covers PyTorch, TensorFlow, JAX, scikit-learn, boosting libraries, ONNX, Hugging Face, and related production tools.
Can it design both batch and real-time inference?
Yes. It can compare batch, streaming, online, cloud, and edge serving patterns using workload requirements.
Does it deploy or operate infrastructure?
No. It provides designs, implementation guidance, checks, and runbooks, but execution requires approved tools and environment access.
Can it help with model monitoring?
Yes. It can define data quality, drift, performance, latency, error, business, and alerting metrics.
Can it recommend a cloud platform?
Yes. Provide architecture, team skills, budget, compliance, and portability needs for a useful comparison.
What information should a request include?
Include the use case, data shape, traffic, latency, quality targets, platform, budget, security constraints, and success criteria.