mcp-deploy-manage-agents

mcp-deploy-manage-agents is a skill for Claude Code, Codex from boshi-xixixi/TraeSkill. It costs 22 tokens per session (2,052 once invoked), scanned A, original, MIT.

A guide for deploying and administering MCP-based declarative agents in Microsoft 365, including their sharing, approval, and organizational distribution.

In plain words
What is it for?
Use it to publish agents, assign them to users, manage different agent types, and apply approval and governance processes.
Why use it?
It helps administrators control who can use these agents and how they are governed across an organization.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one. Also seen: installed under .agents/ (shared by several agents).

Good fit Use it to publish agents, assign them to users, manage different agent types, and apply approval and governance processes.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/boshi-xixixi/traeskill/mcp-deploy-manage-agents
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 boshi-xixixi/TraeSkill --skill mcp-deploy-manage-agents
Clone the repo
git clone --depth 1 https://github.com/boshi-xixixi/TraeSkill

Made for: Claude Code, Codex.

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 mcp-deploy-manage-agents

README.md
[![agentmods](https://agentmods.dev/badge/skills/boshi-xixixi/traeskill/mcp-deploy-manage-agents/github.svg)](https://agentmods.dev/skills/boshi-xixixi/traeskill/mcp-deploy-manage-agents)
Your own site
<a href="https://agentmods.dev/skills/boshi-xixixi/traeskill/mcp-deploy-manage-agents"><img src="https://agentmods.dev/badge/skills/boshi-xixixi/traeskill/mcp-deploy-manage-agents/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.

agentmods 80×15 button for mcp-deploy-manage-agents

Your own site · 80×15
<a href="https://agentmods.dev/skills/boshi-xixixi/traeskill/mcp-deploy-manage-agents"><img src="https://agentmods.dev/badge/skills/boshi-xixixi/traeskill/mcp-deploy-manage-agents.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 22 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,052 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.
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 5.1 $0.00022 $0.02052
Opus 5 $0.00011 $0.01026
Sonnet 5 $0.00004 $0.00410
Haiku 4.5 $0.00002 $0.00205

Measured 9d ago against content hash 9da69b1360ab, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-12, from the pricing page.

Security

Grade A, and why

mcp-deploy-manage-agents 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.

.trae/Skills/.agents/skills/mcp-deploy-manage-agents/SKILL.md · 342 lines

How it starts

The opening of the file, as written. The whole thing — 342 lines — stays where its author put it; the contents beside it link to each section on GitHub.

---
mode: 'agent'
tools: ['changes', 'search/codebase', 'edit/editFiles', 'problems']
description: 'Deploy and manage MCP-based declarative agents in Microsoft 365 admin center with governance, assignments, and organizational distribution'
model: 'gpt-4.1'
tags: [mcp, m365-copilot, deployment, admin, agent-management, governance]
---

# Deploy and Manage MCP-Based Agents

Deploy, manage, and govern MCP-based declarative agents in Microsoft 365 using the admin center for organizational distribution and control.

## Agent Types

### Published by Organization
- Built with predefined instructions and actions
- Follow structured logic for predictable tasks
- Require admin approval and publishing process
- Support compliance and governance requirements

### Shared by Creator
- Created in Microsoft 365 Copilot Studio or Agent Builder
- Shared directly with specific users
- Enhanced functionality with search, actions, connectors, APIs
- Visible to admins in agent registry

### Microsoft Agents
- Developed and maintained by Microsoft
- Integrated with Microsoft 365 services
- Pre-approved and ready to use

### External Partner Agents
- Created by verified external developers/vendors
- Subject to admin approval and control
- Configurable availability and permissions

### Frontier Agents
- Experimental or advanced capabilities
- May require limited rollout or additional oversight
- Examples:
  - **App Builder agent**: Managed via M365 Copilot or Power Platform admin center
  - **Workflows agent**: Flow automation managed via Power Platform admin center

## Admin Roles and Permissions

### Required Roles
- **AI Admin**: Full agent management capabilities
- **Global Reader**: View-only access (no editing)

### Best Practices
- Use roles with fewest permissions
- Limit Global Administrator to emergency scenarios
- Follow principle of least privilege

## Agent Management in Microsoft 365 Admin Center

### Access Agent Management
1. Go to [Microsoft 365 admin center](https://admin.microsoft.com/)
2. Navigate to **Agents** page
3. View available, deployed, or blocked agents

### Available Actions

**View Agents**
- Filter by availability (available, deployed, blocked)
- Search for specific agents
- View agent details (name, creator, date, host products, status)

**Deploy Agents**
Options for distribution:
1. **Agent Store**: Submit to Partner Center for validation and public availability
2. **Organization Deployment**: IT admin deploys to all or selected employees

**Manage Agent Lifecycle**
- **Publish**: Make agent available to organization
- **Deploy**: Assign to specific users or groups
- **Block**: Prevent agent from being used
- **Remove**: Delete agent from organization

**Configure Access**
- Set availability for specific user groups
- Manage permissions per agent
- Control which agents appear in Copilot

## Deployment Workflows

### Publish to Organization

**For Agent Developers:**
1. Build agent with Microsoft 365 Agents Toolkit
2. Test thoroughly in development
3. Submit agent for approval
4. Wait for admin review

**For Admins:**
1. Review submitted agent in admin center
2. Validate compliance and security
3. Approve for organizational use
4. Configure deployment settings
5. Publish to selected users or organization-wide

### Deploy via Agent Store

**Developer Steps:**
1. Complete agent development and testing
2. Package agent for submission
3. Submit to Partner Center
4. Await validation process
5. Receive approval notification
6. Agent appears in Copilot store

**Admin Steps:**
1. Discover agents in Copilot store
2. Review agent details and permissions
3. Assign to organization or user groups
4. Monitor usage and feedback

### Deploy Organizational Agent

**Admin Deployment Options:**
```
Organization-wide:
- All employees with Copilot license
- Automatically available in Copilot

Group-based:
- Specific departments or teams
- Security group assignments
- Role-based access control
```

**Configuration Steps:**
1. Navigate to Agents page in admin center
2. Select agent to deploy
3. Choose deployment scope:
   - All users
   - Specific security groups
   - Individual users
4. Set availability status
5. Configure permissions if applicable
6. Deploy and monitor

## User Experience

### Agent Discovery
Users find agents in:
- Microsoft 365 Copilot hub
- Agent picker in Copilot interface
- Organization's agent catalog

### Agent Access Control
Users can:
- Toggle agents on/off during interactions
- Add/remove agents from their experience
- Right-click agents to manage preferences
- Only access admin-allowed agents

### Agent Usage
- Agents appear in Copilot sidebar
- Users select agent for context
- Queries routed through selected agent
- Responses leverage agent's capabilities

## Governance and Compliance

### Security Considerations
- **Data access**: Review what data agent can access
- **API permissions**: Validate required scopes
- **Authentication**: Ensure secure OAuth flows
- **External connections**: Assess risk of external integrations

### Compliance Requirements
- **Data residency**: Verify data stays within boundaries
- **Privacy policies**: Review agent privacy statement
- **Terms of use**: Validate acceptable use policies
- **Audit logs**: Monitor agent usage and activity

### Monitoring and Reporting
Track:
- Agent adoption rates
- User feedback and satisfaction
- Error rates and performance
- Security incidents or violations

## MCP-Specific Management

### MCP Agent Characteristics
- Connect to external systems via Model Context Protocol
- Use tools exposed by MCP servers
- Require OAuth 2.0 or SSO authentication
- Support same governance as REST API agents

### MCP Agent Validation
Verify:
- MCP server URL is accessible
- Authentication configuration is secure
- Tools imported are appropriate
- Response data doesn't expose sensitive info
- Server follows security best practices

### MCP Agent Deployment
Same process as REST API agents:
1. Review in admin center
2. Validate MCP server compliance
3. Test authentication flow
4. Deploy to users/groups
5. Monitor performance

## Agent Settings and Configuration

### Organizational Settings
Configure at tenant level:
- Enable/disable agent creation
- Set default permissions
- Configure approval workflows
- Define compliance policies

### Per-Agent Settings
Configure for individual agents:
- Availability (on/off)
- User assignment (all/groups/individuals)
- Permission scopes
- Usage limits or quotas

### Environment Routing
For Power Platform-based agents:
- Configure default environment
- Enable environment routing for Copilot Studio
- Manage flows via Power Platform admin center

## Shared Agent Management

### View Shared Agents
Admins can see:
- List of all shared agents
- Creator information
- Creation date
- Host products
- Availability status

### Manage Shared Agents
Admin actions:
- Search for specific shared agents
- View agent capabilities
- Block unsafe or non-compliant agents
- Monitor agent lifecycle

### User Access to Shared Agents
Users access through:
- Microsoft 365 Copilot on various surfaces
- Agent-specific tasks and assistance
- Creator-defined capabilities

## Best Practices

### Before Deployment
- **Pilot test** with small user group
- **Gather feedback** from early adopters
- **Validate security** and compliance
- **Document** agent capabilities and limitations
- **Train users** on agent usage

### During Deployment
- **Phased rollout** to manage adoption
- **Monitor performance** and errors
- **Collect feedback** continuously
- **Address issues** promptly
- **Communicate** availability to users

### Post-Deployment
- **Track metrics**: Adoption, satisfaction, errors
- **Iterate**: Improve based on feedback
- **Update**: Keep agent current with new features
- **Retire**: Remove obsolete or unused agents
- **Review**: Regular security and compliance audits

### Communication
- Announce new agents to users
- Provide documentation and examples
- Share best practices and use cases
- Highlight benefits and capabilities
- Offer support channels

## Troubleshooting

### Agent Not Appearing
- Check deployment status in admin center
- Verify user is in assigned group
- Confirm agent is not blocked
- Check user has Copilot license
- Refresh Copilot interface

### Authentication Failures
- Verify OAuth credentials are valid
- Check user has necessary permissions
- Confirm MCP server is accessible
- Test authentication flow independently

### Performance Issues
- Monitor MCP server response times
- Check network connectivity
- Review error logs in admin center
- Validate agent isn't rate-limited

### Compliance Violations
- Block agent immediately if unsafe
- Review audit logs for violations
- Investigate data access patterns
- Update policies to prevent recurrence

## Resources

- [Microsoft 365 admin center](https://admin.microsoft.com/)
- [Power Platform admin center](https://admin.powerplatform.microsoft.com/)
- [Partner Center](https://partner.microsoft.com/) for agent submissions
- [Microsoft Agent 365 Overview](https://learn.microsoft.com/en-us/microsoft-agent-365/overview)
- [Agent Registry Documentation](https://learn.microsoft.com/en-us/microsoft-365/admin/manage/agent-registry)

## Workflow

Ask the user:
1. Is this agent ready for deployment or still in development?
2. Who should have access (all users, specific groups, individuals)?
3. Are there compliance or security requirements to address?
4. Should this be published to the organization or the public store?
5. What monitoring and reporting is needed?

Then provide:
- Step-by-step deployment guide
- Admin center configuration steps
- User assignment recommendations
- Governance and compliance checklist
- Monitoring and reporting plan

Read the full file on GitHub · 342 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. 9d ago First seen · 342 lines · 22 tokens per session scan A 9da69b1360ab

Subscribe to this mod's changes

mcp-deploy-manage-agents is a skill published in the GitHub repository boshi-xixixi/TraeSkill (263 stars, last pushed 4mo ago), licensed MIT. It adds 22 tokens to every session and 2,052 once invoked, about $0.0001 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 2026-09-03.

Related

Other skills, from other repositories

to-issues

A planning aid that breaks a PRD or technical specification into small implementation issues, each covering a complete, demonstrable piece of work.

smallnest/goal-workflow · 96 tokens

fable-orchestrate

Use when the user wants a task decomposed and driven by a Fable 5.1 orchestrator delegating implementation to Sonnet 5 workers. Fable decomposes into self-contained specs, dispatches Sonnet workers (Agents for 1–2 pieces, a Workflow for fan-out), reviews each result, merges into one branch, and takes a binding verdict…

richkuo/rk-skills · 115 tokens

fable-validate-loop

Use when the user asks to validate a GitHub issue with Fable 5.1 and then autonomously drive it to a reviewed PR in one shot — "fable-validate-loop", "fable validate and work on.

richkuo/rk-skills · 52 tokens

fable-validate-fableplan-loop

Use when the user asks to validate a GitHub issue with Fable 5.1, always plan it with Fable 5.1, and autonomously drive it to a reviewed PR in one shot — "fable-validate-fableplan-loop", "fable validate, fable plan, and work on.

richkuo/rk-skills · 74 tokens

milestoneplan

Use when the user wants a milestone's execution plan shown as a table — "milestoneplan v1", "/milestoneplan", "show the plan for v1". Read-only - reads every issue in the milestone and renders one table of issue number, description, complexity, dependencies, validate/build model and effort, fableplan, plan effort, and…

richkuo/rk-skills · 100 tokens

new-issue

Use when the user says "new issue", "create an issue", "file an issue", "/new-issue", or asks to turn a bug, idea, or discussion into a GitHub issue. Takes an optional description of what the issue should cover; with no input, derives the issue from what was discussed in the current conversation. Verifies claims…

richkuo/rk-skills · 94 tokens