aiskillstore/marketplace / skills/89jobrien/ai-ethics
ai-ethics
VerifiedResponsible AI development and ethical considerations. Use when evaluating AI bias, implementing fairness measures, conducting ethical assessments, or ensuring AI systems align with human values.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 5.7 KB |
| skill-report.json | config | 10.3 KB |
skill-report.json
View on GitHub{
"schema_version": "2.0",
"meta": {
"generated_at": "2026-07-05T11:10:14.084Z",
"slug": "89jobrien-ai-ethics",
"source_url": "https://github.com/89jobrien/steve/tree/main/steve/skills/ai-ethics",
"source_ref": "main",
"model": "codex",
"analysis_version": "3.0.0",
"source_type": "community",
"content_hash": "05ef6d78162c33414ed5eafb33d915267ac81d3411428745dff79c19d0e2ec6e",
"tree_hash": "82eaf0671c0589b1f00ef2e39a0b91f0edf418e2eabc6a570b4e1f18d35d97d9",
"provenance": {
"model_effective": "codex",
"fallback_chain": [
{
"agent": "codex",
"outcome": "success"
}
],
"analysis_version": "3.0.0"
}
},
"skill": {
"name": "ai-ethics",
"description": "Evaluate AI bias, fairness, transparency, governance, privacy, and regulatory compliance for responsible AI development.",
"summary": "Guides responsible AI reviews, bias mitigation, oversight, and compliance planning.",
"icon": "📦",
"version": "1.0.2",
"author": "Joseph OBrien",
"license": "MIT",
"tags": [
"ai-ethics",
"bias-assessment",
"fairness",
"governance",
"compliance"
],
"supported_tools": [
"claude",
"codex",
"claude-code"
],
"risk_factors": [
"external_commands"
],
"category": "research"
},
"security_audit": {
"risk_level": "safe",
"is_blocked": false,
"safe_to_publish": true,
"analysis_status": "ok",
"summary": "The two static findings are false positives caused by Markdown code-formatted reference file names, not shell execution. No prompt injection, data exfiltration intent, or executable behavior was found in SKILL.md.",
"remediation": [],
"risk_factor_evidence": [
{
"factor": "external_commands",
"evidence": [
{
"file": "SKILL.md",
"line_start": 201,
"line_end": 201
},
{
"file": "SKILL.md",
"line_start": 202,
"line_end": 202
}
]
}
],
"static_findings": [
{
"id": "external_commands:SKILL.md:201:ruby-shell-backtick-execution",
"category": "external_commands",
"severity": "medium",
"pattern": "Ruby/shell backtick execution",
"file": "SKILL.md",
"line_start": 201,
"line_end": 201,
"snippet": "- **`references/bias_assessment.md`** - Detailed bias evaluation methodology"
},
{
"id": "external_commands:SKILL.md:202:ruby-shell-backtick-execution",
"category": "external_commands",
"severity": "medium",
"pattern": "Ruby/shell backtick execution",
"file": "SKILL.md",
"line_start": 202,
"line_end": 202,
"snippet": "- **`references/regulatory_compliance.md`** - AI regulation requirements"
}
],
"finding_verdicts": [
{
"id": "external_commands:SKILL.md:201:ruby-shell-backtick-execution",
"verdict": "false_positive",
"confidence": 0.97,
"reason": "The snippet is a Markdown bullet with a backtick-wrapped reference file path, not Ruby or shell execution. It does not execute commands or pass user input to a shell."
},
{
"id": "external_commands:SKILL.md:202:ruby-shell-backtick-execution",
"verdict": "false_positive",
"confidence": 0.97,
"reason": "The snippet is a Markdown bullet with a backtick-wrapped reference file path, not Ruby or shell execution. It does not execute commands or pass user input to a shell."
}
],
"semantic_findings": [],
"critical_findings": [],
"high_findings": [],
"medium_findings": [],
"low_findings": [],
"dangerous_patterns": [],
"files_scanned": 1,
"total_lines": 209,
"audit_model": "codex",
"audited_at": "2026-07-05T11:10:14.084Z",
"agent_auto_install_policy": "allowed",
"manual_install_policy": "allowed"
},
"content": {
"user_title": "Evaluate AI Ethics Risk",
"value_statement": "AI teams must identify bias, oversight gaps, and compliance risks before deployment. This skill guides structured reviews for fairness, transparency, privacy, governance, and responsible AI documentation.",
"seo_keywords": [
"AI ethics skill",
"Claude AI ethics",
"Codex AI ethics",
"Claude Code AI governance",
"responsible AI",
"bias assessment",
"fairness metrics",
"AI governance",
"model cards",
"EU AI Act compliance"
],
"actual_capabilities": [
"Identifies common AI bias types and mitigation strategies.",
"Summarizes group and individual fairness metrics.",
"Outlines model card documentation fields and explainability methods.",
"Maps AI systems to governance and oversight patterns.",
"Highlights privacy and environmental impact considerations.",
"Summarizes EU AI Act risk categories."
],
"limitations": [
"Does not run model audits or compute fairness metrics itself.",
"Does not provide legal advice or guarantee regulatory compliance.",
"Requires accurate system, data, and deployment context from the user.",
"Does not include executable scripts or automated testing tools."
],
"use_cases": [
{
"title": "Plan Ethical Product Reviews",
"description": "Assess user impact, transparency needs, oversight points, and documentation gaps before launch.",
"target_user": "AI product managers"
},
{
"title": "Choose Fairness Checks",
"description": "Compare bias sources, subgroup metrics, mitigation options, and monitoring needs for a model.",
"target_user": "Machine learning teams"
},
{
"title": "Prepare Governance Evidence",
"description": "Organize risk classification, model cards, audit trails, and human oversight records for review.",
"target_user": "Compliance reviewers"
}
],
"prompt_templates": [
{
"title": "Check Basic AI Ethics Risks",
"scenario": "Beginner ethics review",
"prompt": "Review this AI system concept for fairness, transparency, privacy, accountability, safety, and human agency. System: [describe system]."
},
{
"title": "Create a Bias Assessment Plan",
"scenario": "Model fairness planning",
"prompt": "Build a bias assessment plan for this model. Include likely bias sources, subgroup checks, fairness metrics, and mitigation options. Context: [model and data]."
},
{
"title": "Draft Model Card Content",
"scenario": "Responsible AI documentation",
"prompt": "Draft model card sections for this AI system. Include intended use, training data summary, subgroup performance needs, limitations, and ethical considerations. System: [details]."
},
{
"title": "Design Governance for High Risk AI",
"scenario": "Advanced governance workflow",
"prompt": "Design an AI governance workflow for this high impact system. Include risk classification, review gates, human oversight, audit trails, incident response, and compliance evidence. Context: [use case, users, region]."
}
],
"output_examples": [
{
"input": "Review a hiring screening model for ethical risks.",
"output": [
"The review flags historical and representation bias as primary concerns.",
"Recommended checks include subgroup selection rates, false negative rates, and appeal outcomes.",
"Human oversight should include review paths for rejected candidates and audit logs."
]
},
{
"input": "Create model card guidance for a medical triage assistant.",
"output": [
"The model card should define intended clinical support use and prohibited autonomous decisions.",
"It should report subgroup performance, data provenance, limitations, and escalation requirements.",
"The documentation should state monitoring plans and incident response ownership."
]
},
{
"input": "Assess governance needs for a customer credit model.",
"output": [
"The system should be treated as high impact because decisions affect financial access.",
"Governance evidence should include risk approval, explanation procedures, audit trails, and complaint handling.",
"Periodic bias monitoring should compare outcomes across protected and proxy groups."
]
}
],
"best_practices": [
"Provide concrete details about users, affected groups, data sources, and deployment context.",
"Ask for subgroup metrics and impact analysis when decisions affect people.",
"Use the skill before deployment and again after monitoring reveals new risks."
],
"anti_patterns": [
"Using only aggregate accuracy to claim an AI system is fair.",
"Treating model explanations as a replacement for human oversight.",
"Requesting legal conclusions without providing jurisdiction, system risk level, or evidence."
],
"faq": [
{
"question": "Can this skill run fairness tests?",
"answer": "No. It helps plan and interpret reviews, but it does not execute tests or inspect datasets."
},
{
"question": "Does it cover the EU AI Act?",
"answer": "Yes. It summarizes EU AI Act risk categories and documentation needs. It is not legal advice."
},
{
"question": "Which AI systems benefit most?",
"answer": "Systems that affect access, safety, finances, employment, healthcare, education, or public services benefit most."
},
{
"question": "Can it help create model cards?",
"answer": "Yes. It outlines model purpose, data, subgroup metrics, limitations, and ethical considerations for model cards."
},
{
"question": "Does it replace an ethics board?",
"answer": "No. It supports structured preparation and review, but human governance remains necessary."
},
{
"question": "Can Claude, Codex, and Claude Code use it?",
"answer": "Yes. The skill declares support for Claude, Codex, and Claude Code."
}
]
},
"file_structure": [
{
"name": "SKILL.md",
"type": "file",
"path": "SKILL.md",
"lines": 209
}
]
}
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |