Agent QA Gates

A field-tested validation system for AI agent output. Born from production failures, not theory.

Quick Start

Before any agent delivers output, run the Pre-Ship Checklist:

1. Accurate? — every number/date/metric has a source. Unsourced → prefix "estimated"
2. Complete? — no missing pieces, no "I'll do that next"
3. Actionable? — ends with clear next step or decision point
4. Fits the channel? — check character limits for your delivery surface
5. No leaks? — no internal context, private data, or secrets
6. Not a duplicate? — verify no recent identical send
7. Would the human be embarrassed? — if yes, don't ship

Gate Tiers

Four ascending tiers by risk level: