Based on Reddit community insight: "The hardest part isn't orchestration.
It's alignment." This module computes a composite alignment score from
truthfulness, instruction compliance, safety, and behavioral consistency.
The Alignment Index is a single number (0-1) that answers:
"How well does this agent do what it's supposed to do, safely?"
Alignment Index — Composite trust signal
Based on Reddit community insight: "The hardest part isn't orchestration. It's alignment." This module computes a composite alignment score from truthfulness, instruction compliance, safety, and behavioral consistency.
The Alignment Index is a single number (0-1) that answers: "How well does this agent do what it's supposed to do, safely?"