Observability & Evaluation · Data artifact

Reference Trajectory

Data artifactObservability & EvaluationObservability & Evaluationarc:ReferenceTrajectory

A ground-truth expected action sequence for a test case, listing each tool, its expected parameters and outputs, required ordering, context conditions and success criteria.

Responsibility. Specifies the correct action path against which agent trajectories are scored.

Also known as: Golden Trajectory, Expected Action Sequence

Direct neighbourhood (hover for relationship types)

Relationships

configures structural

Trajectory Matching Evaluator Ch3.8

is read by dependency

Design guidance

SHOULD annotate reference actions with context conditions (e.g., user requested a human) so justified deviations are credited.
SHOULD specify security-critical parameter sources (e.g., user_id from authentication context).

Classification

Patterns: Context-conditioned reference actions

Sources

Ch3.8: T. Nguyen, "Action Accuracy Metrics," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 3.8. ISBN: 9798244538229.