| name | specify-jira |
| description | Specify skill for Jira ticket analysis — defines analysis categories, skills to load, agents to launch, and severity classification |
Spec Filename Prefix
jira- (followed by ticket ID, e.g., jira-BW-10257)
Skills to Load
None required (domain skills loaded based on project type).
Agents to Launch
None specified.
Analysis Categories
Prerequisites
- Verify
acli is installed: command -v acli
- Extract ticket ID from branch name using pattern
[A-Z]+-[0-9]+
Ticket Fetching
acli jira workitem view <TICKET-ID> --fields "summary,description,status,priority,issuelinks,comment,attachment"- If ticket has linked issues, fetch those for additional context
Requirement Analysis
- Parse ticket summary, description, and comments to extract full scope
- Identify acceptance criteria, constraints, and technical details
- If description is vague or ambiguous, ask for clarification
Codebase Mapping
- Identify which files, modules, and layers the work touches
- Determine new code vs existing code modifications
- Identify inputs, outputs, and side effects
- Note dependencies between tasks
Spec Output Sections
- Ticket: Key, summary, status, priority, Jira link
- Overview: 2-3 sentence summary of what to build and why
- Architecture: How changes fit into existing codebase
- Tasks: Ordered list with file path, changes needed, dependencies, complexity, parallelizability
- API contracts: New endpoints, function signatures, type definitions
- Edge cases: Known edge cases, error conditions, boundary behaviors
- Testing approach: What tests are needed and what they verify
- Open questions: Ambiguities grouped by requirements, architecture, scope, risks
Severity Classification
Not applicable — this is implementation planning, not issue finding.
Scope Overrides
Scope is always derived from the current branch's Jira ticket. If no ticket ID can be extracted from the branch name, notify and stop.
