m365-agent-evaluator

Official

Evaluate and improve Microsoft 365 Copilot declarative agents

Authormicrosoft
Version1.0.0
Installs0

System Documentation

What problem does it solve?

This Skill provides the ability to evaluate and fine-tune Microsoft 365 Copilot declarative agents, enabling users to ensure their agents meet the desired standards of relevance, coherence, and groundedness.

Core Features & Use Cases

  • Evaluation: Use the CLI to run evaluation suites on agents and analyze results.
  • Dataset Creation: Design schema-compatible eval datasets for specific agent behaviors.
  • Result Analysis: Analyze output for relevance, coherence, groundedness, and similarity scores.
  • Iterative Improvement: Apply fixes based on analysis to improve agent performance.
  • Use Case: A developer can use this Skill to evaluate and optimize their Copilot agent's performance, ensuring it meets the required quality standards for a production environment.

Quick Start

Use the m365-agent-evaluator skill to evaluate your Microsoft 365 Copilot agent with the following command:

npx -y --package @microsoft/m365-copilot-eval@latest runevals --prompts-file evals\evals.json --output .evals\results.json

Dependency Matrix

Required Modules

@microsoft/m365-copilot-evalnode

Components

scriptsreferencesassets

💻 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: m365-agent-evaluator
Download link: https://github.com/microsoft/skills/archive/main.zip#m365-agent-evaluator

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 620,000+ vetted skills library on demand.