copilot-studio
Skill46 starsChanged 8 months agoMITView on GitHub
Microsoft Copilot Studio agent building. Use when: creating Copilot agents,
In this repository40 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- accessibility-uxSkill
- ai-builderSkill
- almSkill
- architectureSkill
- azure-openaiSkill
- business-process-flowsSkill
- canvas-appsSkill
- code-appsSkill
- code-reviewerSkill
- custom-connectorsSkill
- dashboardsSkill
- data-migrationSkill
- dataverse-mcpSkill
- dataverseSkill
- dataverse-web-apiSkill
- devops-engineerSkill
- env-strategySkill
- governanceSkill
- integration-patternsSkill
- licensingSkill
- m365-integrationSkill
- model-driven-appsSkill
- observabilitySkill
- pcfSkill
- perf-optimiseSkill
- platform-builderSkill
- pluginsSkill
- power-automateSkill
- power-biSkill
- power-pagesSkill
- power-platform-teamSkill
- project-managerSkill
- securitySkill
- solutions-architectSkill
- spec-driven-devSkill
- test-engineSkill
- testingSkill
- uat-coordinatorSkill
- web-resourcesSkill
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 4 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈4
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 47
No fallback topic configured (agent goes silent on unrecognized input)
- Distinctive: line 49
No escalation path to a human agent
- Distinctive: line 52
No adversarial/prompt injection testing before deployment
- Distinctive: line 55
No analytics monitoring after deployment (can't measure agent effectiveness)
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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.
Files
Skill: Microsoft Copilot Studio
When to Use
Trigger when building, configuring, testing, deploying, or reviewing Copilot Studio agents (formerly Power Virtual Agents) — including standalone agents, Copilot extensions, autonomous agents, and multi-agent orchestrations.
Agent Types
| Type | Description | Use Case |
| Copilot agent | Custom agent with topics + generative AI | Customer support, employee self-service |
| Copilot extension | Extends Microsoft 365 Copilot | Enterprise Copilot with custom knowledge |
| Autonomous agent | Event-driven, runs without user interaction | Monitoring, proactive notifications |
| Connected agent | Agent-to-agent via A2A protocol | Cross-domain orchestration |
| Embedded agent | Agent embedded within another agent | Specialized sub-agent for delegation |
⚠️ REQUIRED: Load Sub-Files Before Implementation
SKILL.md is a summary only — it is NOT sufficient for implementation.
The detailed content (complete payloads, XML templates, working examples, edge-case handling) lives in sub-files in the same directory as this SKILL.md. Before writing any code, you MUST use read_file on the sub-files relevant to your task:
- Topics & Conversation Authoring — topic types, trigger phrases, conversation flow design, variables, conditions, topic management
- Knowledge Sources — Dataverse, SharePoint, websites, files, Graph, configuration, content moderation, grounding
- Multi-Agent Orchestration — embedded agents, connected agents, A2A protocol, agent delegation, cross-domain patterns
- Autonomous Agents — Agent Flows in Power Automate, generative actions, event-driven triggers, Automation Center governance, computer use (Wave 1 2026)
- MCP Integration — MCP as tool pattern in Copilot Studio, connector infrastructure (VNet, DLP, auth), when MCP vs connector actions
- Channels & Deployment — Teams, website, mobile, Omnichannel, Agent API for Power Pages, authentication per channel
- ALM, Governance & Testing — solution packaging, environment promotion, adversarial testing, analytics, Copilot tuning
Anti-Patterns
- No fallback topic configured (agent goes silent on unrecognized input)
- Generative answers enabled without content moderation
- No escalation path to a human agent
- Hardcoded values in topic actions instead of environment variables
- Agents outside of solutions (can't be promoted via ALM)
- No adversarial/prompt injection testing before deployment
- Using authentication-required actions without configuring agent authentication
- Overly broad knowledge sources (agent responds to off-topic queries)
- No analytics monitoring after deployment (can't measure agent effectiveness)
- Using MCP tools for schema creation instead of direct Web API (MCP is for reads/queries)
- Multi-agent orchestration without defined security boundaries between agents
- Autonomous agents without human-in-the-loop gates for destructive actions
Related Skills
power-automate— Agent Flows, plugin actions via flowsazure-openai— BYOM, AI architecture decisionsdataverse— Structured knowledge sourcessecurity— Agent authentication configurationalm— Solution packaging for agent promotionpower-pages— Agent API for custom chat experiences
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
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.

