bmad-github-story-create

Story Create: Sync + BMAD Create-Story

You are creating a new story for development. This skill orchestrates two phases:

1. Sync local state (lightweight refresh)
2. Run the BMAD create-story operation (creates the story file with full user engagement)

After the operation, the user reviews the output, commits when happy, then runs Story Dev (SD) to start implementation.

Worktree behavior

This skill writes a new file into <output_folder>/implementation-artifacts/ on whichever branch is currently checked out. It can run from main or from any worktree, but be aware:

• From main (recommended for the standard sequential flow): the new story file is committed on main, and any subsequent /story-dev worktree branches from a main that already has the file.
• From a worktree on some other branch (e.g., Conductor spawned you in a planning worktree): the new story file lands on that branch. You must merge it to main before another agent's /story-dev can see it.

If you're unsure where you are, run git branch --show-current and decide based on the branch name.

---

Phase 0: Sync First

Read and follow ${CLAUDE_PLUGIN_ROOT}/references/sync-first.md.

---

Phase 1: Run BMAD create-story

Read and follow ${CLAUDE_PLUGIN_ROOT}/references/bmad-workflow-loader.md with <operation> = create-story.

After the operation completes, note:

• The story ID (e.g., 1-2)
• The story key/slug from the story file name (e.g., 1-2-bridge-interface-and-shared-type-contracts)
• The story title (e.g., Bridge Interface & Shared Type Contracts)
• The story file path in <output_folder>/implementation-artifacts/

If BMAD halts because sprint-status.yaml is missing (this can happen if the project has never been sprint-planned, or you're on a branch where the file hasn't been merged yet), tell the user to either run BMAD sprint-planning or pass an explicit <epic>-<story> identifier (e.g., 1-2) to create-story to bypass auto-discovery.