What is analyze-evals?

A tool for reading CSV evaluation results exported from Microsoft Copilot Studio, a service for building AI agents. It finds failed tests and suggests fixes in YAML, a text format for configuration.

Use it to review an agent's test questions, compare expected and actual answers, examine failure explanations, and draft YAML changes for problems it finds.

It turns a results file into a focused list of failures, so you do not have to inspect every test response manually.

analyze-evals — Skill for Claude Code by microsoft

About the project

Skills for Copilot Studio is a plugin for authoring, testing, and troubleshooting standard Microsoft Copilot Studio agents as YAML files from a terminal or editor. It is intended for users of Claude Code, GitHub Copilot CLI, and VS Code who work with Copilot Studio agents. The catalogue add-ons are the plugin's skills, agents, hook, and plugin definition.

microsoft/skills-for-copilot-studio · 433 stars · on GitHub

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent

npx skills add microsoft/skills-for-copilot-studio --skill analyze-evals

Clone the repo

git clone --depth 1 https://github.com/microsoft/skills-for-copilot-studio

Made for: Claude Code.

Or install copilot-studio, the plugin that ships this one along with the rest of its 31 skills, 4 agents, 1 hook.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

README.md

[![agentmods](https://agentmods.dev/badge/skills/microsoft/skills-for-copilot-studio/analyze-evals/github.svg)](https://agentmods.dev/skills/microsoft/skills-for-copilot-studio/analyze-evals)

Your own site

<a href="https://agentmods.dev/skills/microsoft/skills-for-copilot-studio/analyze-evals"><img src="https://agentmods.dev/badge/skills/microsoft/skills-for-copilot-studio/analyze-evals/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/skills/microsoft/skills-for-copilot-studio/analyze-evals"><img src="https://agentmods.dev/badge/skills/microsoft/skills-for-copilot-studio/analyze-evals.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 52 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 757 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits

NVIDIA SkillSpector pass 7 Sept 2026

How audits are shown

Origin original No closer match found in the catalogue.

Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

Model	Per session	Once invoked
Fable 5.1	$0.00052	$0.00757
Opus 5	$0.00026	$0.00378
Sonnet 5	$0.00010	$0.00151
Haiku 4.5	$0.00005	$0.00076

Measured 9d ago against content hash 7768f4f3dc20, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-08, from the pricing page.

Security

Grade A, and why

analyze-evals scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 9d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

skills/analyze-evals/SKILL.md · 63 lines

How it starts

The opening of the file, as written. The whole thing — 63 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Analyze Copilot Studio Evaluation Results

Analyze evaluation results exported from the Copilot Studio UI as CSV.

Phase 1: Get Results

Ask the user for the CSV file path if not already provided. The file is typically exported from Copilot Studio's Evaluate tab and named Evaluate <agent name> <date>.csv in their Downloads folder.

Read the CSV file. The in-product evaluation CSV has these columns:

Column	Meaning
`question`	The test utterance
`expectedResponse`	Expected response (may be empty)
`actualResponse`	What the agent responded
`testMethodType_1`	Eval method (e.g., `GeneralQuality`)
`result_1`	`Pass` or `Fail`
`passingScore_1`	Score threshold (may be empty)
`explanation_1`	Why it passed/failed (e.g., "Seems relevant; Seems incomplete; Knowledge sources not cited")

The _1 suffix indicates the first eval method. There may be additional methods (_2, _3, etc.) with the same column pattern.

Phase 2: Analyze Results

Focus on failed evaluations (result_1 = Fail, or any result_N = Fail).
For each failure, use the explanation column to understand the issue:
- "Question not answered" — The agent couldn't handle the question. Check if there's a matching topic or knowledge source.
- "Knowledge sources not cited" — The agent responded but didn't cite sources. Check knowledge source configuration and SearchAndSummarizeContent nodes.
- "Seems incomplete" — The response was partial. Check topic flow for early exits, missing branches, or incomplete SendActivity messages.
- Error messages in actualResponse (e.g., GenAIToolPlannerRateLimitReached) — These are runtime errors, not authoring issues. Flag them to the user as transient failures to retry.

Phase 3: Propose Fixes

For each failure, identify the relevant YAML file(s):
- Auto-discover the agent: Glob: **/agent.mcs.yml
- Find the relevant topic by matching the test utterance against trigger phrases and model descriptions
- Read the topic file to understand the current flow

Read the full file on GitHub · 63 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

9d ago First seen · 63 lines · 52 tokens per session scan A 7768f4f3dc20

Subscribe to this mod's changes

analyze-evals is a skill published in the GitHub repository microsoft/skills-for-copilot-studio (433 stars, last pushed 5d ago), licensed MIT. It adds 52 tokens to every session and 757 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Analytics and metrics

systematic-debugging

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.

obra/superpowers · 21 tokens

local-ai-agents

Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…

microsoft/ai-agents-for-beginners · 200 tokens

next-cache-components-adoption

Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…

vercel/next.js · 95 tokens

next-cache-components-optimizer

Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…

vercel/next.js · 170 tokens

next-partial-prefetching-adoption

Turn on Partial Prefetching in a Next.js app and work through the insights it surfaces. Use when the user wants to enable or adopt Partial Prefetching, flip the partialPrefetching flag, opt routes in with export const prefetch = 'partial', audit Link prefetch={true} behavior, preserve existing prefetched UI with…

vercel/next.js · 103 tokens

chronicle

Analyze Copilot session history for standup reports, usage tips, session search, and session reindexing. Use when the user asks for a standup, daily summary, usage tips, workflow recommendations, wants to search or find past sessions by keyword/file/PR, wants to reindex their session store, or asks about deleting…

microsoft/vscode · 72 tokens