repo-native-alignment: Skill for Claude Code

jira-plan

Creates well-structured Jira subtasks ready for autonomous agents. Works in two modes:

1. Task mode: Investigates codebase, asks clarifying questions, creates subtasks
2. Session mode: Reads existing .oh/<session>.md context, skips investigation, creates subtasks

Prerequisites

This skill requires a Jira MCP server to be configured. At startup, verify the MCP is available:

Check for Jira MCP tools:
- jira_search or jira_search_issues
- jira_create_issue
- jira_get_issue

If tools are not available:
  → Exit with: "Jira MCP not configured. Add a Jira MCP server to use this skill."

Invocation

# Task mode - investigate from scratch
/jira-plan PROJ-123 "Add dark mode support to the dashboard"

# Session mode - use existing context from skills workflow
/jira-plan PROJ-123 auth-refactor

Arguments:

• First argument: Parent issue key (e.g., PROJ-123) - subtasks will be created under this
• Second argument: Task description OR session name

Mode Detection

At the start, check if the second argument matches an existing session file:

If .oh/<arg>.md exists AND contains ## Solution Space:
    → Session mode (skip investigation, use context)
Else:
    → Task mode (investigate, ask questions)

Session Mode Flow

When .oh/<session>.md exists with solution-space context:

1. Read session file and extract:

   • ## Aim → Goal for subtasks
   • ## Problem Space → Constraints, terrain, assumptions
   • ## Solution Space → Selected approach, trade-offs accepted
   • ## Problem Statement → Framing (if present)

2. Skip investigation - context already gathered by grounding skills

3. Skip clarifying questions - decisions already made during solution-space

4. Decompose the selected solution into right-sized subtasks:

   • Use the Solution Space recommendation as the starting point
   • Apply standard decomposition heuristics (see below)
   • Aim for 1-4 hours of work per subtask