create-jira-tickets

name	create-jira-tickets
description	Create Jira Epic + Stories for a list of planned tasks.
allowed-tools	[]
required-mcps	["jira"]
inputs	{"tasks":{"type":"array","required":true},"project_key":{"type":"string","required":true},"epic_title":{"type":"string","required":true}}
outputs	{"tickets":{"type":"array"},"epic":{"type":"object"},"errors":{"type":"array"}}
llm	{"max_iterations":30,"temperature":0.1}

create-jira-tickets

Create a Jira Epic + Stories for each task, return the created ticket references.

Inputs

• tasks (object[], required): output of plan-tasks.
• project_key (string, required): Jira project key (e.g. "ABC").
• epic_title (string, required): title for the umbrella epic.

Instructions

• Use the jira MCP server's tools. Typical calls: create_issue with issuetype.
• Create ONE Epic first, capture its key, then create all stories with customfield_* = Epic link or parent reference (whichever the MCP supports).
• If a Jira API call returns an error, record it in errors rather than retrying blindly.
• Preserve the original planned task data in every returned ticket object. The downstream implementer consumes tickets directly, so each ticket must keep the task id, title, description, acceptance_criteria, depends_on, and any delivery_profile exactly as received.
• Add Jira metadata to the preserved task object instead of replacing the task with a thin reference. Use task_id for the original task id and include ticket_key plus url when available.

Outputs

• tickets (object[]): original task objects enriched with {task_id, ticket_key, url}
• epic (object): {key, url}
• errors (string[])