Getting started

Skill obielin/responsible-ai-skills/skills/getting-started

Skills framework for coding agents that enforces responsible AI practices — bias assessment, fairness testing, explainability, governance documentation, and alignment review. Auto-activates when building AI systems.

Install

npx -y skills add obielin/responsible-ai-skills --skill getting-started

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

2 stars2 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Use when starting any AI/ML task, or when asked about responsible AI skills. Explains available skills and mandatory workflow.

SKILL.md

2.4 KB, as published. Nobody here has run it

Responsible AI Skills — Getting Started

You have a suite of Responsible AI Skills. These are mandatory workflows, not suggestions.

What These Skills Do

They ensure every AI system you build is:

Explainable — decisions can be understood and justified
Fair — performance is equitable across population groups
Governed — documentation exists for audit and accountability
Aligned — behaviour matches stated intent and human values
Safe — incidents have a response plan before they happen

Available Skills

Skill	When It Activates
`bias-assessment`	Loading datasets, training models, evaluating model performance
`fairness-testing`	Writing tests for any ML model or classifier
`explainability-by-default`	Building prediction, classification, or recommendation systems
`governance-documentation`	Before deploying any AI system to production
`responsible-data-handling`	Accessing, loading, or processing datasets
`human-oversight-design`	Designing autonomous or agentic AI systems
`ai-incident-response`	AI system behaves unexpectedly, produces harmful output, or underperforms
`alignment-review`	Before marking any AI feature complete

Mandatory Rules

Before building any AI feature — check if a skill applies. If it does, read and follow it.
Before calling any AI task "done" — run alignment-review.
Before deploying to production — run governance-documentation.
You cannot skip these steps by claiming time pressure or simplicity.

How Skills Work

Skills load progressively — this file uses ~100 tokens. Full skill content loads only when relevant. This means you can have all skills available without context cost.

Quick Reference

Building a classifier?        → bias-assessment + fairness-testing + explainability-by-default
Loading training data?        → responsible-data-handling
Deploying to production?      → governance-documentation
Designing an agent?           → human-oversight-design
Something went wrong?         → ai-incident-response
Finishing any AI feature?     → alignment-review (ALWAYS)

Read the relevant skill now. Do not proceed with your task until you have.