# Jamf MCP Server - Claude Skills
This directory contains pre-built Claude skills for common Jamf management tasks. These skills provide high-level workflows that combine multiple MCP tools to accomplish complex operations.
## 📁 Directory Structure
```
skills/
├── device-management/ # Device-focused operations
├── policy-management/ # Policy deployment and management
├── automation/ # Scheduled tasks and compliance
├── reporting/ # Analytics and reporting (coming soon)
└── types.ts # Common type definitions
```
## 🛠️ Available Skills
### Device Management
#### `find-outdated-devices`
Identify devices that haven't checked in recently.
**Example usage:**
```typescript
await findOutdatedDevices(context, {
daysSinceLastContact: 30,
includeDetails: true
});
```
#### `batch-inventory-update`
Update inventory for multiple devices efficiently.
**Example usage:**
```typescript
await batchInventoryUpdate(context, {
deviceIdentifiers: ['C02XL1234567', 'C02XL2345678'],
identifierType: 'serialNumber',
maxConcurrent: 5
});
```
### Policy Management
#### `deploy-policy-by-criteria`
Deploy policies to devices based on specific criteria like OS version, department, or check-in status.
**Example usage:**
```typescript
await deployPolicyByCriteria(context, {
policyIdentifier: 'Security Update',
identifierType: 'name',
criteria: {
osVersion: '14',
daysSinceLastContact: 7
},
dryRun: true
});
```
### Automation
#### `scheduled-compliance-check`
Perform comprehensive compliance audits with detailed reporting.
**Example usage:**
```typescript
await scheduledComplianceCheck(context, {
checks: {
outdatedDevices: { enabled: true, daysThreshold: 30 },
osVersionCompliance: { enabled: true, minimumVersion: '14.0' },
missingProfiles: {
enabled: true,
requiredProfiles: ['FileVault', 'Firewall']
}
},
outputFormat: 'detailed'
});
```
## 🚀 Using Skills in Claude
### Basic Pattern
When using these skills with Claude, follow this pattern:
1. **Identify the task** - Determine which skill matches your need
2. **Gather parameters** - Collect necessary information (device IDs, policy names, etc.)
3. **Run with dry-run** - Test with `dryRun: true` first
4. **Execute** - Run with actual parameters and confirmation
### Example Conversations
**Finding outdated devices:**
```
User: "Show me all devices that haven't checked in for 2 weeks"
Claude: I'll use the find-outdated-devices skill to identify devices that haven't checked in for 14 days...
```
**Deploying policies:**
```
User: "Deploy the security update policy to all Engineering Macs running macOS 13"
Claude: I'll use the deploy-policy-by-criteria skill to target specific devices. Let me first do a dry run to show you which devices would be affected...
```
**Compliance checking:**
```
User: "Run a compliance check for outdated devices and OS versions"
Claude: I'll perform a comprehensive compliance check using the scheduled-compliance-check skill...
```
## 💡 Best Practices
### 1. Always Start with Dry Runs
Most skills support a `dryRun` parameter. Use it to preview actions before execution.
### 2. Use Specific Identifiers
When possible, use specific identifiers (IDs, serial numbers) rather than names for accuracy.
### 3. Batch Operations
For operations on multiple devices, use batch skills to avoid rate limiting and improve performance.
### 4. Check Results
Always review the skill's output, especially the `success` field and any error messages.
### 5. Follow Up Actions
Many skills provide `nextActions` suggestions. Use these to guide follow-up tasks.
## 🔧 Creating Custom Skills
To create a new skill:
1. Choose the appropriate category directory
2. Create a new TypeScript file following the naming pattern
3. Implement the skill function and metadata
4. Use the types from `types.ts`
5. Add comprehensive error handling
6. Include usage examples in metadata
### Skill Template
```typescript
import { SkillContext, SkillResult } from '../types';
interface YourSkillParams {
// Define parameters
}
export async function yourSkillName(
context: SkillContext,
params: YourSkillParams
): Promise<SkillResult> {
try {
// Implementation
return {
success: true,
message: 'Operation completed',
data: { /* results */ }
};
} catch (error) {
return {
success: false,
message: `Operation failed: ${error.message}`,
error
};
}
}
export const metadata = {
name: 'your-skill-name',
description: 'What this skill does',
parameters: {
// Define parameter metadata
},
examples: [
// Provide examples
]
};
```
## 📊 Skill Categories
### Device Management
- Device discovery and inventory
- Batch operations
- Health monitoring
### Policy Management
- Policy deployment
- Configuration management
- Software distribution
### Automation
- Scheduled tasks
- Compliance checking
- Automated remediation
### Reporting (Coming Soon)
- Analytics dashboards
- Executive summaries
- Trend analysis
## 🔗 Integration with MCP Tools
Skills are built on top of the core MCP tools:
- `searchDevices` - Find devices
- `checkDeviceCompliance` - Check device status
- `executePolicy` - Deploy policies
- `updateInventory` - Refresh device data
- And more...
Skills combine these tools to create powerful workflows while handling errors, validating inputs, and formatting outputs for better usability.
