eval-guide is an AI agent evaluation toolkit for Copilot Studio that helps users plan evaluations, create test cases, interpret results, and diagnose failures. It is used with Claude Code or GitHub Copilot to assess single-response and multi-turn agent behavior using Microsoft’s evaluation guidance.
Borrowing it
Nothing to install: this file belongs to microsoft/eval-guide. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/microsoft/eval-guide/main/.github/copilot-instructions.mdgit clone --depth 1 https://github.com/microsoft/eval-guideWrote 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.
[](https://agentmods.dev/instructions/microsoft/eval-guide/copilot-instructions)<a href="https://agentmods.dev/instructions/microsoft/eval-guide/copilot-instructions"><img src="https://agentmods.dev/badge/instructions/microsoft/eval-guide/copilot-instructions/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.
<a href="https://agentmods.dev/instructions/microsoft/eval-guide/copilot-instructions"><img src="https://agentmods.dev/badge/instructions/microsoft/eval-guide/copilot-instructions.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 | $0.00422 | $0.00422 |
| Opus 5.5 | $0.00169 | $0.00169 |
| Sonnet 5.5 | $0.00084 | $0.00084 |
| Haiku 4.5 | $0.00042 | $0.00042 |
Grade A, and why
eval-guide copilot-instructions.md 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 2mo 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.
What it actually says
Eval Guide — Copilot Instructions
This repository is an AI agent evaluation toolkit for Copilot Studio. It provides structured methodologies for planning, generating, running, and interpreting AI agent evaluations.
Quick Reference
- Prompt files are in
.github/prompts/— attach the relevant one when helping with eval tasks - Full methodology is the 10-step playbook in
skills/eval-guide/playbook.md(canonical spine);AGENTS.mdat the repo root is the cross-tool summary - Eval runner script is at
skills/eval-guide/scripts/eval-runner.js
Skill Routing
| Task | Prompt File |
|---|---|
| Full eval lifecycle | eval-guide.prompt.md |
| Create eval plan | eval-suite-planner.prompt.md |
| Generate test cases | eval-generator.prompt.md |
| Interpret results | eval-result-interpreter.prompt.md |
| Triage failures | eval-triage-and-improvement.prompt.md |
| Methodology Q&A | eval-faq.prompt.md |
Key Conventions
- CSV format for Copilot Studio import: exactly 2 columns —
Question,Expected response. The testing method is assigned per row in the Copilot Studio Evaluate tab after import (not a CSV column); other methodology metadata lives in the companion.docxmanifest - Valid test methods (assigned in the UI):
General quality,Compare meaning,Text similarity,Exact match,Keyword match(core five), plusCapability useandCustom(extensions) - Always explain reasoning — users should learn the methodology, not just receive artifacts
- For planner output, prefer the interactive HTML review page over long chat summaries; keep chat to artifact paths and blockers
- Include at least 1 adversarial/safety eval set or case in every eval plan
- Group test cases by eval set into separate CSV files
- Stages 0-2 (Discover, Plan, Generate) work without a running agent
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.
- 2mo ago First seen · 31 lines · 422 tokens per session scan A a3450652182c
eval-guide copilot-instructions.md is an instructions file published in the GitHub repository microsoft/eval-guide (140 stars, last pushed 4mo ago), licensed MIT. It adds 422 tokens to every session, about $0.0017 per session on Opus 5.5. 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.
Other instructions, from other repositories
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
next.js AGENTS.md
AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, cache eligibility and rollout and applying the notice (cutover).
langchain AGENTS.md
AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.
vscode copilot-instructions.md
Copilot instructions for microsoft/vscode, covering vs code copilot instructions, project overview, root folders, core architecture (src/ folder) and built-in extensions (extensions/ folder).
deepseek-harness AGENTS.md
AGENTS.md instructions for deepseek-ai/deepseek-harness, covering agents.md, pre-stable apis and released session data, repository layout, commands and host sandbox failures.