Trajectory Evaluation

Community

Evaluate multi-turn agent trajectories for quality.

Authorreaatech
Version1.0.0
Installs0

System Documentation

What problem does it solve?

Trajectory evaluation assesses the quality of complete agent executions—multi-turn conversations with tool invocations and task completion—by measuring coherence, goal completion, and turn-level quality.

Core Features & Use Cases

  • Holistic evaluation of conversation flow and outcome.
  • Coherence assessment, goal completion tracking, and turn-level analysis.
  • Regression detection across releases and comparisons against golden trajectories.
  • Use cases include validating agent performance in multi-turn tasks and surfacing quality issues for tuning.

Quick Start

Evaluate a trajectory file with the CLI or programmatic API to generate a comprehensive quality report and per-turn diagnostics.

Dependency Matrix

Required Modules

None required

Components

Standard package

đź’» Claude Code Installation

Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.

Please help me install this Skill:
Name: Trajectory Evaluation
Download link: https://github.com/reaatech/agent-eval-harness/archive/main.zip#trajectory-evaluation

Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
View Source Repository

Agent Skills Search Helper

Install a tiny helper to your Agent, search and equip skill from 471,000+ vetted skills library on demand.