jira-assistant

The Jira Assistant is an expert-level agent skill designed for seamless integration between your coding environment and Atlassian Jira. It leverages the Atlassian MCP (Model Context Protocol) to allow AI agents to act directly within your project management workflow. This skill is intended for software engineers, project leads, and product managers who need to maintain synchronization between their codebase and Jira tracking without switching contexts. By centralizing issue management, it ensures that technical documentation and development tasks remain aligned.

• Advanced Jira search capabilities using both Rovo natural language queries and precision-based JQL filtering for tasks, bugs, and epics.

• Automated configuration detection by parsing .cursor/rules/jira-config.mdc, identifying your specific Cloud ID, Project Key, and Board URL to prevent operational errors.

• Full support for the Jira issue lifecycle, including creating new issues with enforced markdown templates, updating field metadata, and performing workflow transitions.

• Intelligent metadata retrieval, such as fetching project-specific issue types and required fields, ensuring every created ticket meets organizational quality standards.

• Efficient sprint management, enabling users to move tickets, update statuses like 'To Do' to 'In Progress' or 'Done', and add team comments directly.

• Always begin by using the search tool to explore existing issues before creating duplicates.

• When creating tickets, strictly follow the provided task template, which includes sections for Context, Objective, Technical Requirements, and Acceptance Criteria to ensure high-quality documentation.

• Always include project = {PROJECT_KEY} in JQL queries to restrict scope and improve performance.

• Note that this skill is strictly for Jira management; for Confluence-related operations, use the dedicated confluence-assistant to prevent configuration conflicts.

• If workspace configuration is missing, the skill will prompt the user to discover available projects via MCP or specify a project key manually.

• Avoid hardcoding file paths in issue descriptions, as codebase structure evolves over time; instead, focus on high-level technical objectives and requirements.