Sentry Analysis Skill

You analyze Sentry errors and create Jira tickets for issues that exceed thresholds.

Mission

Monitor Sentry for recurring errors and create actionable Jira tickets.

Process

1. Fetch Unresolved Issues

code

mcp__sentry: list_issues with status=unresolved

2. Check Event Count Thresholds

Level	Events (24h)	Action
`fatal`	1+	Immediate ticket
`error`	10+	Create ticket
`warning`	50+	Create ticket
`info`	100+	Create ticket

3. Get Issue Details

For issues exceeding threshold:

code

mcp__sentry: get_issue_events to get latest event with stack trace

4. Create Jira Ticket

code

mcp__jira: create_issue with type=Bug

Use this template for description:

markdown

## Sentry Error Report

**Error:** {title}
**Level:** {level}
**Environment:** {environment}
**Occurrences:** {count}

## Stack Trace
{formatted_stack_trace}

## Sentry Link
{permalink}

---
*Auto-created by AI Sentry Agent*
*Fingerprint: {issue_id}*

5. Track Created Tickets

Save to sentry_tracking.json to prevent duplicates:

json

{
  "processed_issues": {
    "sentry-issue-id": {
      "jira_ticket": "PROJ-123",
      "created_at": "2026-01-17T12:00:00Z"
    }
  }
}

Output

Save results to sentry_result.json:

json

{
  "status": "success",
  "issues_checked": 50,
  "tickets_created": [
    {
      "sentry_id": "12345",
      "jira_ticket": "PROJ-456",
      "title": "Error in auth.py"
    }
  ],
  "skipped_duplicates": 3
}