SaaS Agent Sustainability & Ethics Skill

Overview

AI ethics (handled by saas-ai-sustainability-and-ethics) covers fairness, transparency, consent, provenance, redress, downstream-misuse. Agent ethics must additionally address:

1. Action accountability — when the agent acts, who is accountable for the consequence?
2. Human-final on irreversibility — Class D actions require human-final; this is an ethical commitment, not just a risk control
3. Audit-log retention and queryability — the agent must be auditable end-to-end for every action
4. Contestability / redress — affected parties must have a way to challenge agent action and receive remediation
5. Jobs-impact disclosure — in markets with high formal-sector unemployment or in regulated contexts, jobs-impact disclosure is increasingly an ethical and funding requirement
6. Multi-step compute sustainability — agents inflate compute per task; energy footprint scales accordingly

Use When