MicrosoftKnown publishermicrosoft/skills-for-copilot-studio/skills/edit-agent/SKILL.md

edit-agent

Skill486 starsChanged 8 months agoMITView on GitHub

Edit Copilot Studio agent settings, instructions, or configuration. Use when the user asks to change agent instructions, display name, conversation starters, AI settings, or generative actions toggle.

In this repository30 other files it gives its agents.
- add-actionSkill
- add-adaptive-cardSkill
- add-generative-answersSkill
- add-global-variableSkill
- add-knowledgeSkill
- add-nodeSkill
- add-other-agentsSkill
- analyze-evalsSkill
- chat-directlineSkill
- chat-sdkSkill
- chat-with-agentSkill
- clone-agentSkill
- create-eval-setSkill
- create-evalSkill
- detect-modeSkill
- directline-chatSkill
- edit-actionSkill
- edit-triggersSkill
- int-patternsSkill
- int-project-contextSkill
- int-referenceSkill
- list-kindsSkill
- list-topicsSkill
- lookup-schemaSkill
- manage-agentSkill
- new-topicSkill
- run-evalSkill
- run-tests-kitSkill
- test-authSkill
- validateSkill

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 3 are distinctive, and 2 are very specific to this project.

≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.

Common
Distinctive
Very specific

Distinctive:
NEVER hardcode an agent name.
line 20
Very specific:
Use the output to determine the correct modelNameHint and provider values.
line 30
Very specific:
CRITICAL: Do NOT include kind in the model block — the agent-level schema uses CurrentModelsNoKind.
line 31
Distinctive:
Read the current file: before making any changes.
line 33
Distinctive:
If the user asks to change any of these, warn them.
line 69

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.

Files

SKILL.md

Edit Agent Settings/Instructions

Modify agent metadata (agent.mcs.yml) or configuration (settings.mcs.yml).

Instructions

Auto-discover the agent directory:
```
Glob: **/agent.mcs.yml
```
NEVER hardcode an agent name. If multiple agents found, ask which one.
Identify what the user wants to change and read the appropriate file:
- Instructions, display name, conversation starters, AI settings → agent.mcs.yml
- GenerativeActionsEnabled, authentication, recognizer, capabilities → settings.mcs.yml
If the user wants to change the AI model, run the schema lookup tool first:
```
node ${CLAUDE_SKILL_DIR}/../../scripts/schema-lookup.bundle.js models
```
Use the output to determine the correct modelNameHint and provider values. CRITICAL: Do NOT include kind in the model block — the agent-level schema uses CurrentModelsNoKind.
Read the current file before making any changes.
Make the requested changes using the Edit tool.

Editable Fields in `agent.mcs.yml`

Field	Description	Example
`displayName`	Agent's display name	`displayName: My Agent`
`instructions`	System prompt / personality	Multi-line YAML with `\|`
`conversationStarters`	Suggested conversation starters	Array of `{title, text}`
`aISettings.model.modelNameHint`	Model hint (run `models` command for convention)	`Sonnet46`, `GPT5Chat`
`aISettings.model.provider`	Model provider (required for non-OpenAI)	`Anthropic`

Editable Fields in `settings.mcs.yml`

Field	Description	Example
`GenerativeActionsEnabled`	Enable generative orchestration	`true` / `false`
`authenticationMode`	Auth mode	`Integrated`, `None`
`authenticationTrigger`	When to auth	`Always`, `AsNeeded`
`accessControlPolicy`	Access control	`ChatbotReaders`
`configuration.aISettings.*`	AI capabilities	Various booleans
`configuration.settings.*.content.capabilities.webBrowsing`	Web browsing	`true` / `false`

Writing Effective Instructions

For detailed guidance on writing knowledge-aware instructions, grounding directives, citation control, and scope enforcement, see instructions-guide.md.

Fields to NEVER Modify

schemaName — This is the internal Power Platform identifier. Changing it breaks the agent.
publishedOn — Managed by the platform.
template — Managed by the platform.
language — Changing this can corrupt the agent.

If the user asks to change any of these, warn them.

Example: Update Instructions

instructions: |
  You are a customer support agent for Contoso Ltd.

  Guidelines:
  - Be professional and empathetic
  - Always verify the customer's identity first
  - Escalate billing issues to a human agent

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

It worked
Partly
It didn't

What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free