Jira Cloud Skill Pack Files

id: jira_prepare_issue_fields
title: "Prepare Issue Fields"
description: "Converts unstructured element content into structured Jira issue fields using LLM"
overview: |
  A pre-processing primitive that takes a raw block of text — usually the contents of a highlighted DOM element, a clipboard paste, or a stack trace — and maps it to Jira's expected field shape: summary, description, priority, labels, and component. An LLM step does the extraction; the output is a structured object ready for a create call.

  This workflow does not create the ticket itself. Compose it in front of a create step when the source content is messy or mixed-format and needs normalising before submission.

category:
  level: primitive
  domain: support

params:
  element_text: string
  page_url: string
  page_title: string

steps:
  - type: llm.generate
    prompt: |
      Convert this webpage element into a Jira issue summary and description.

      Element content:
      {{element_text}}

      Page: {{page_title}} ({{page_url}})

      Output ONLY valid JSON (no markdown, no explanation):
      {
        "summary": "Brief title, max 100 chars",
        "description": "Detailed description with context"
      }

      Rules:
      - Summary: concise, specific, actionable (e.g. "Review Fireflies HubSpot integration guide")
      - Description: include source page URL and relevant details from the element
    as: issue_fields_json

  - type: variable.set
    name: issue_fields
    value: "{{issue_fields_json}}"