agent-eval-harness

Official

Structured AI agent evaluation pipelines.

Authorplaited
Version1.0.0
Installs0

System Documentation

What problem does it solve?

This Skill provides a CLI-driven framework to capture agent trajectories, apply schema-driven adapters, and emit standardized JSONL outputs for downstream scoring and evaluation.

Core Features & Use Cases

  • Trajectory capture: Run prompts against headless adapters to collect full trajectories, including thoughts, tools, and plans, in JSONL.
  • Grading and analysis: Calibrate graders, summarize outputs, and compare configurations across runs.
  • Pipeline-friendly: Integrates into CI, supports multiple commands (capture, extract, grade, summarize, calibrate, balance, compare).

Quick Start

Use the agent-eval-harness to capture prompts with a headless adapter and produce results.jsonl for analysis.

Dependency Matrix

Required Modules

None required

Components

Standard package

💻 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: agent-eval-harness
Download link: https://github.com/plaited/agent-eval-harness/archive/main.zip#agent-eval-harness

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 223,000+ vetted skills library on demand.