Borrowing it
Nothing to install: this file belongs to DuqueOM/ML-MLOps-Portfolio. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/DuqueOM/ML-MLOps-Portfolio/main/.devin/skills/new-service/SKILL.mdgit clone --depth 1 https://github.com/DuqueOM/ML-MLOps-PortfolioWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/duqueom/ml-mlops-portfolio/new-service)<a href="https://agentmods.dev/skills/duqueom/ml-mlops-portfolio/new-service"><img src="https://agentmods.dev/badge/skills/duqueom/ml-mlops-portfolio/new-service/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/duqueom/ml-mlops-portfolio/new-service"><img src="https://agentmods.dev/badge/skills/duqueom/ml-mlops-portfolio/new-service.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00017 | $0.02243 |
| Opus 5 | $0.00009 | $0.01122 |
| Sonnet 5 | $0.00003 | $0.00449 |
| Haiku 4.5 | $0.00002 | $0.00224 |
Grade C, and why
new-service scanned grade C with 2 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 9d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Recursive force deletehighDestructive command
rm -rf with a variable or a broad path is one typo away from removing the wrong tree.
scaffold_files: AUTO # reversible by `rm -rf <ServiceName>/` Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
**Success criteria**: `pytest tests/test_fastapi_template_contract.py tests/test_api.py -v` passes. `curl localhost:8000/health` returns healthy and `/ready` returns 200 only after the model is loaded and warmed. How it starts
The opening of the file, as written. The whole thing — 229 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Create New ML Service
Guides creation of a complete, production-ready ML service using the template system.
Inputs
$service-name: Service slug (e.g.,bankchurn,frauddetect)$business-problem: What the service predicts/classifies
Goal
A fully deployed, tested, monitored ML service with all quality gates passing, drift detection running, and documentation complete.
Pre-conditions
templates/scripts/new-service.shexists and is executable- The caller has specified ServiceName (PascalCase) and service_slug (snake_case)
- Cloud target is known (gcp, aws, or both)
Steps
1. Gather Requirements
Human checkpoint: Confirm requirements before scaffolding.
Answer these questions:
- Business problem: What does this service predict/classify/estimate?
- Dataset: Source, size, features, target distribution
- Model type: Classification, regression, NLP, time series?
- Scale: Expected request volume, latency requirements
- Explainability: Is SHAP required? (High-stakes decisions = yes)
2. Run Scaffolding Script
bash templates/scripts/new-service.sh "$service-name" "$service-slug"
Verify no remaining placeholders:
grep -r "{ServiceName}\|{service}\|{SERVICE}" $service-name/ --include="*.py" --include="*.yaml" | head -20
Success criteria: Directory created with zero remaining {ServiceName}, {service}, or {SERVICE} placeholders. Run examples/minimal/ if this is the first time to validate template works.
3. Data Validation (Agent-DataValidator)
- Define Pandera schema in
src/$service-name/schemas.py - Check for temporal data → review for leakage risk
- Create background data for SHAP (50 representative samples)
- Version data with DVC:
dvc add data/raw/dataset.csv
Success criteria: Pandera schema validates sample data without errors. DVC tracking configured.
4. Training Pipeline (Agent-MLTrainer)
- Implement
FeatureEngineerclass insrc/$service-name/training/features.py - Define model pipeline in
src/$service-name/training/model.py - Implement
Trainer.run()insrc/$service-name/training/train.py:- load_data() + Pandera validation
- engineer_features()
- split_train_val_test() (temporal if dates exist)
- cross_validate() with StratifiedKFold
- evaluate() with optimal threshold
- fairness_check() (DIR >= 0.80)
- save_artifacts() with SHA256
- log_to_mlflow()
- quality_gates()
- Configure Optuna (minimum 50 trials)
- Create MLflow experiment
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 9d ago First seen · 229 lines · 17 tokens per session scan C 2ea94a9ffdd5
new-service is a skill published in the GitHub repository DuqueOM/ML-MLOps-Portfolio (5 stars, last pushed 2d ago), licensed MIT. It adds 17 tokens to every session and 2,243 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it C with 2 findings (recursive force delete, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
mlflow
Track ML experiments, manage model registry with versioning, deploy models to production, and reproduce experiments with MLflow - framework-agnostic ML lifecycle platform.
setup
Configure MLflow tracing for Claude Code.
prowler-mcp
Creates MCP tools for Prowler MCP Server. Covers BaseTool pattern, model design, and API client usage. Trigger: When working in mcpserver/ on tools (BaseTool), models (MinimalSerializerMixin/fromapiresponse), or API client patterns.
dstack-prototyping
Use with the dstack skill for model-serving work when the image, serving command, resources, backend/fleet choice, or service behavior is not proven. Guides task-first prototyping on real hardware, choosing fleets/backends that can reuse idle instances and caches, checking vLLM/SGLang sources, and verifying the final…
claude-api
Build, debug, and optimize Claude API / Anthropic SDK apps. Apps built with this skill should include prompt caching. Also handles migrating existing Claude API code between Claude model versions (4.5 → 4.6, 4.6 → 4.7, retired-model replacements). TRIGGER when: code imports anthropic/@anthropic-ai/sdk; user asks for…
calling-llms
Use when sending chat completions through liter-llm and routing to a specific provider via the provider/model prefix. Covers the chat call shape, provider routing, modelhint, message roles, and error categories.