From powerball-workplace
Fetch and analyze Jira tickets to produce actionable development plans. Use this skill whenever the user provides a Jira ticket number (e.g., XY-12345), a Jira URL, or asks to analyze/plan work for a Jira issue. Also trigger when the user says things like 'what needs to be done for [ticket]', 'plan the work for [ticket]', 'analyze [ticket]', 'look at this Jira ticket', or mentions working on a specific ticket. This skill fetches the ticket via the Atlassian MCP server, analyzes the problem, searches the codebase for relevant files, and produces a structured development plan including reproduction steps (for bugs), implementation steps, and test coverage guidance.
npx claudepluginhub tim-hub/powerball --plugin powerball-workplaceThis skill uses the workspace's default tool permissions.
Fetch a Jira ticket, understand the problem, explore relevant code, and produce a development plan with testing guidance.
Fetches JIRA issues by key or search, distills into structured tasks, analyzes codebase for gaps and risks, optionally enriches JIRA or spawns subtasks. Use before coding workflows.
Views Jira issues, generates git branches from tickets, creates tickets, transitions status via Atlassian MCP. Activates on ticket keys (XX-123), /jira, branch/ticket creation requests.
Reads and analyzes JIRA issues from Red Hat Issue Tracker: fetches details/comments/custom fields/Git links, searches via JQL, categorizes bugs/CVEs/features/stories, traverses ticket graphs (parents/children/links), generates summaries. Read-only Python tool.
Share bugs, ideas, or general feedback.
Fetch a Jira ticket, understand the problem, explore relevant code, and produce a development plan with testing guidance.
Requires the Atlassian MCP server to be connected. If tools like getJiraIssue are unavailable, ask the user to run /mcp to connect and authenticate.
If the user has configured their cloudId in CLAUDE.md, use it directly instead of calling getAccessibleAtlassianResources.
The user may provide a ticket number (XY-12345), a Jira URL, or a description. For descriptions, search with searchJiraIssuesUsingJql and ask the user to confirm which ticket.
Fetch with getJiraIssue(cloudId, issueIdOrKey, responseContentFormat="markdown") and extract: summary, type, status, priority, description, acceptance criteria, comments, linked issues, attachments, and labels/components.
Fetch up to 3 linked issues (parent epic, blockers, related tickets, confluence pages) for broader context. Pay special attention to comments — they often contain the most valuable context (decisions, workarounds, reproduction details).
Classify the ticket type — this drives the plan structure:
Synthesize context from description, comments, linked tickets, and labels to write a 2-3 sentence problem summary for a developer who has never seen this ticket.
Search for relevant code using the ticket context:
Grep for keywords (error messages, component names, function/class names)Glob for file patterns (e.g., **/Calendar/**/*.php)Read to understand the code once locatedStart narrow — if the ticket mentions specific files or classes, begin there. Optionally run git log --oneline -10 <file> on critical files to check for recent activity or ongoing work.
Output using this structure, adapting sections based on ticket type.