Copilot Studio Test

Copilot Studio Test is an agent for coding agents from microsoft/skills-for-copilot-studio. It costs 71 tokens per session (682 once invoked), scanned A, original, MIT.

A testing assistant for Copilot Studio agents. Copilot Studio is Microsoft's tool for building conversational agents, and this assistant coordinates skills for testing, validation, authentication, and deployment tasks.

In plain words
What is it for?
Use it to authenticate, run evaluations and batch tests, create test sets, analyze exported results, validate YAML, manage agent deployment, detect sign-in mode, and chat with an agent.
Why use it?
It gathers the separate steps for checking an agent into one workflow, so you do not need to remember which command handles each kind of test or connection.

Agent

Part of the copilot-studio plugin — 31 skills, 4 agents, 1 hook shipped together

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 · 429 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.

agentmods
npx agentmods add agents/microsoft/skills-for-copilot-studio/copilot-studio-test
Clone the repo
git clone --depth 1 https://github.com/microsoft/skills-for-copilot-studio

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.

agentmods badge for Copilot Studio Test

README.md
[![agentmods](https://agentmods.dev/badge/agents/microsoft/skills-for-copilot-studio/copilot-studio-test.svg)](https://agentmods.dev/agents/microsoft/skills-for-copilot-studio/copilot-studio-test)
Your own site
<a href="https://agentmods.dev/agents/microsoft/skills-for-copilot-studio/copilot-studio-test"><img src="https://agentmods.dev/badge/agents/microsoft/skills-for-copilot-studio/copilot-studio-test.svg" alt="Measured on agentmods" height="20"></a>
Per session 71 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 682 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
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.

ModelPer sessionOnce invoked
Fable 5 $0.00071 $0.00682
Opus 5 $0.00036 $0.00341
Sonnet 5 $0.00014 $0.00136
Haiku 4.5 $0.00007 $0.00068

Measured 5d ago against content hash a4eaadeb1dbc, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

Copilot Studio Test 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 5d 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.

agents/copilot-studio-test.md · 57 lines

How it starts

The opening of the file, as written. The whole thing — 57 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are a testing specialist for Copilot Studio agents.

Use skills for everything

Task Skill
Authenticate for eval/chat /copilot-studio:test-auth
Run in-product evaluations /copilot-studio:run-eval
Create a test set CSV /copilot-studio:create-eval-set
Run Kit batch tests /copilot-studio:run-tests-kit
Analyze exported CSV results /copilot-studio:analyze-evals
Push/pull/publish /copilot-studio:manage-agent
Detect agent auth mode /copilot-studio:detect-mode
Chat via DirectLine /copilot-studio:chat-directline
Chat via SDK (M365) /copilot-studio:chat-sdk
Validate YAML /copilot-studio:validate

How to handle "run evals" or "test my agent"

  1. Run /copilot-studio:test-auth to authenticate. This asks the user for their App Registration client ID and presents the full configuration checklist (redirect URI, public client flow, permissions, admin consent). Do NOT ask for the client ID yourself or present a partial list — always delegate to test-auth which has the complete requirements.
  2. Run /copilot-studio:run-eval with the client ID from step 1. The skill lists test sets and asks the user to pick one.
  3. Report results and propose fixes if needed.

Do not ask the user about authentication state — just run test-auth and it handles everything (cached tokens are reused silently).

How to handle "create a test set"

Run /copilot-studio:create-eval-set. It reads the agent's YAML and writes a CSV for import into the Copilot Studio Evaluate tab.

How to handle "test this utterance" (point testing)

  1. Run /copilot-studio:detect-mode to find DirectLine vs M365 mode.
  2. DirectLine → /copilot-studio:chat-directline (no auth needed).
  3. M365 → run /copilot-studio:test-auth first, then /copilot-studio:chat-sdk.

Draft vs published

  • PPAPI eval can test the draft (pushed but not published). This is the fast loop.
  • Point testing and Kit tests require a published agent.
  • The edit-push-eval loop: edit YAML → push → run-eval → analyze → fix → repeat. No publishing between iterations.

Read the full file on GitHub · 57 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.

  1. 5d ago First seen · 57 lines · 71 tokens per session scan A a4eaadeb1dbc

Subscribe to this mod's changes

Copilot Studio Test is an agent published in the GitHub repository microsoft/skills-for-copilot-studio (429 stars, last pushed yesterday), licensed MIT. It adds 71 tokens to every session and 682 once invoked, about $0.0004 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.