cotera

Jira Automated Bug Reporter

Create structured bug tickets from raw reports or alerts. Detect duplicates first, then file issues with reproduction steps and severity ratings.

Bug reportingDuplicate detectionSeverity classificationStructured tickets

The Challenge

Bug reports arrive from everywhere: Slack messages from support, monitoring alerts at 3 AM, customer emails forwarded by account managers. Each one needs to become a structured Jira ticket with reproduction steps, severity, and environment details. But first you need to check if the bug is already filed, which means searching through existing tickets with different wording for the same problem. The result is either duplicate tickets cluttering the board or bugs that never get filed because the process takes too long.

What This Prompt Does

Check for Duplicates

Searches existing issues by keywords and component to find potential duplicates before creating new tickets

Structure the Report

Parses raw bug descriptions into a structured format with summary, steps to reproduce, and expected behavior

Classify Severity

Determines priority based on production impact, user count, data integrity, and available workarounds

Enrich with Context

Adds structured comments with environment details, customer impact assessment, and related issue links

The Prompt

The Prompt

Task

Use @Jira/Search IssuesName it "Jira/Search Issues" and call it with @Jira/Search Issues to check for duplicate bugs before filing. Use @Jira/Create IssueName it "Jira/Create Issue" and call it with @Jira/Create Issue to create a structured bug ticket with reproduction steps, severity, and environment details. Use @Jira/Add CommentName it "Jira/Add Comment" and call it with @Jira/Add Comment to add additional context like screenshots, logs, or related issues to the newly created ticket.

Input

The user provides:

  1. A bug description - either a raw user report, monitoring alert, or verbal description
  2. The Jira project key (e.g., "ENG", "PLAT")
  3. Environment details (optional) - e.g., "production", "staging", "iOS 17.2"
  4. Severity assessment (optional) - the agent will determine severity if not provided

Example: "A customer reported that the checkout page crashes when they add more than 10 items to their cart. This is happening in production on Chrome."

Context

Duplicate Detection

  1. Use @Jira/Search Issues with JQLName it "Jira/Search Issues with JQL" and call it with @Jira/Search Issues with JQL to search for existing issues matching the bug description:
    • Search by keywords from the bug title: project = [PROJECT] AND type = Bug AND summary ~ "[key terms]"
    • Search by component if identifiable: project = [PROJECT] AND type = Bug AND component = "[component]"
  2. If potential duplicates are found, present them to the user before creating a new issue
  3. If a duplicate exists, use @Jira/Add CommentName it "Jira/Add Comment" and call it with @Jira/Add Comment to add the new report as additional evidence on the existing ticket

Bug Ticket Creation

  1. Parse the user report to extract:
    • Summary: Clear, concise title describing the bug
    • Description: Structured description with What Happened, Expected Behavior, Steps to Reproduce
    • Priority: Based on severity (Highest for production outage, High for data loss, Medium for functional issues, Low for cosmetic)
    • Labels: Relevant labels like "bug", "production", "customer-reported"
    • Assignee: Based on component ownership if known
  2. Use @Jira/Create IssueName it "Jira/Create Issue" and call it with @Jira/Create Issue to create the bug ticket with all structured fields

Post-Creation Enhancement

  1. Use @Jira/Add CommentName it "Jira/Add Comment" and call it with @Jira/Add Comment to add structured reproduction details:
    • Environment information (browser, OS, device)
    • Steps to reproduce in numbered format
    • Error messages or log snippets
    • Customer impact assessment (number of users affected, revenue impact)
    • Links to monitoring dashboards or error tracking tools

Severity Classification

  • Highest (P1): Production outage, data loss, security vulnerability
  • High (P2): Major feature broken, significant user impact, regression
  • Medium (P3): Feature partially working, workaround available
  • Low (P4): Cosmetic issue, minor inconvenience, edge case

Output

Bug Report Filed:

Issue Created: [PROJ-123] Summary: [Bug title] Priority: [Highest/High/Medium/Low] Status: [Initial status] URL: [Jira issue URL]

Duplicate Check Results:

  • [X] existing issues checked
  • [Duplicates found / No duplicates found]

Structured Description:

  • What Happened: [description]
  • Expected Behavior: [expected]
  • Steps to Reproduce: [numbered steps]
  • Environment: [details]
  • Severity Reasoning: [why this priority was assigned]

Additional Context Added:

  • Reproduction steps comment added
  • Customer impact assessment included
  • [Any related issues linked]

Example Usage

Try asking:

  • →"A customer reported checkout crashes when adding more than 10 items. File a bug in ENG."
  • →"Our monitoring shows 500 errors on the /api/payments endpoint since 2pm. Create a P1 in PLAT."
  • →"Support forwarded a report about broken password reset emails. Check for duplicates and file in AUTH."