Start

Start Work on JIRA Ticket

Begin work on JIRA ticket $ARGUMENTS.

Step 0: Session Bootstrap

bash .claude/scripts/session-setup.sh

Handles acli install+auth, recommended hooks, pre-commit, and gh auth check. Runs once per session (skips on subsequent calls).

Step 1: View and Assign

bash .claude/scripts/jira-ops.sh view $ARGUMENTS

Review the ticket summary, status, and description. Then assign and transition:

bash .claude/scripts/jira-ops.sh assign $ARGUMENTS
bash .claude/scripts/jira-ops.sh transition $ARGUMENTS "ASSIGNED"

Step 2: Check Backport Requirement

bash .claude/scripts/jira-ops.sh check-version $ARGUMENTS

If Target Version is set, backporting is required after merge.

Step 3: Create Branch

DEFAULT_BRANCH="${PRIMARY_BRANCH:-$(git rev-parse --abbrev-ref origin/HEAD 2>/dev/null | sed 's|origin/||')}"
DEFAULT_BRANCH="${DEFAULT_BRANCH:-master}"
git checkout "$DEFAULT_BRANCH" && git pull origin "$DEFAULT_BRANCH"
git checkout -b $ARGUMENTS-short-description

Branch naming: <TICKET-KEY>-<kebab-case-description>. Derive the description from the ticket summary. Ask the user if it's ambiguous.

Step 4: Load Context

Based on the ticket's area, read the relevant docs. The doc mapping is project-specific — check the project's AGENTS.md for the area-to-doc table.

Common patterns:

Step 5: Report

Summarize:

• Ticket: key, summary, status, assignee
• Backport: required or not
• Branch: name created
• Docs loaded
• Next step: /code