pipeline-monitor

name: pipeline-monitor description: Monitors Azure DevOps pipeline execution status, retrieves logs, parses errors, and determines success/failure. Polls pipeline runs until completion and provides detailed error analysis. Use for automated pipeline monitoring and debugging.

Pipeline Monitor Skill

Overview

Monitors Azure DevOps pipeline execution, tracks status changes, and analyzes failures. Provides real-time feedback on pipeline progress and detailed error information.

When to Use

• Monitor pipeline execution until completion
• Get current pipeline status without waiting
• Analyze pipeline failures for debugging
• Track pipeline status in automation workflows
• Retrieve detailed failure information for troubleshooting

Prerequisites

• Azure DevOps CLI installed and configured
• Valid Azure authentication (az login)
• Pipeline run ID from pipeline-runner skill
• Access to Azure DevOps organization and project

Available Scripts

1. monitor-pipeline.ps1

Polls pipeline status until completion with progress updates.

Parameters:

• -RunId (required): Pipeline run ID
• -Project (optional): Project name (reads from config if not provided)
• -Organization (optional): Org URL (reads from config if not provided)
• -MaxWaitTime (optional): Timeout in seconds (default: 1800)
• -PollInterval (optional): Polling interval in seconds (default: 10)
• -Quiet (optional): Suppress progress messages

Returns: JSON with runId, status, result, duration, url, succeeded

2. get-pipeline-status.ps1

Quick status check without waiting.

Parameters:

• -RunId (required): Pipeline run ID
• -Project (optional): Project name
• -Organization (optional): Org URL

Returns: JSON with runId, status, result, url, createdDate, finishedDate

3. analyze-failure.ps1

Analyzes failures and provides troubleshooting guidance.

Parameters:

• -RunId (required): Pipeline run ID
• -Project (optional): Project name
• -Organization (optional): Org URL

Returns: JSON with runId, result, url, succeeded, timestamp

Usage Examples

Monitor Pipeline Until Completion

# Basic usage (reads config from starter-config.json)
$result = .\scripts\monitor-pipeline.ps1 -RunId 123 | ConvertFrom-Json

if ($result.succeeded) {
    Write-Host "✅ Pipeline succeeded!"
} else {
    Write-Host "❌ Pipeline failed: $($result.result)"
}

Custom Timeout and Polling

# Wait max 10 minutes, poll every 5 seconds
.\scripts\monitor-pipeline.ps1 -RunId 123 -MaxWaitTime 600 -PollInterval 5

Quick Status Check

# Get current status without waiting
$status = .\scripts\get-pipeline-status.ps1 -RunId 123 | ConvertFrom-Json

Write-Host "Current status: $($status.status)"
Write-Host "Result: $($status.result)"

Analyze Failure

# Get detailed failure analysis
.\scripts\analyze-failure.ps1 -RunId 123

Integrated Workflow

# Run pipeline, monitor, and analyze if failed
$runId = 123  # From pipeline-runner

# Monitor until completion
$result = .\scripts\monitor-pipeline.ps1 -RunId $runId | ConvertFrom-Json

if (-not $result.succeeded) {
    # Analyze failure
    .\scripts\analyze-failure.ps1 -RunId $runId
}

Output Format

All scripts return structured JSON:

{
  "runId": 123,
  "status": "completed",
  "result": "succeeded",
  "duration": 245,
  "url": "https://dev.azure.com/org/project/_build/results?buildId=123",
  "succeeded": true
}

Status values: inProgress, completed, canceling, notStarted, timeout

Result values: succeeded, failed, canceled, partiallySucceeded

Common Error Patterns

Scripts automatically check for these patterns:

Configuration

Scripts automatically read from starter-config.json:

{
  "azureDevOps": {
    "organizationUrl": "https://dev.azure.com/yourorg",
    "projectName": "your-project"
  }
}

You can override with explicit parameters if needed.

Integration Points

• pipeline-runner: Monitor pipelines started by runner
• deployment-validation: Validate pipeline execution results
• pipeline-debugger agent: Automated debugging loops

Best Practices

✅ DO:

• Use default timeout (30 minutes) for most pipelines
• Return structured JSON for downstream processing
• Check both status and result fields
• Include pipeline URL in output
• Handle authentication before calling scripts

❌ DON'T:

• Set very short timeouts (< 5 minutes)
• Poll too frequently (< 5 seconds)
• Ignore the succeeded field in results
• Forget to analyze failures for troubleshooting