菜单
Use when needing to understand a Jira ticket before acting on it — pulling context for an agent, reviewing a ticket's full picture, loading ticket details for decision-making, or when asked to "show me" or "what's in" a ticket.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Use when the user has a spec, architecture doc, implementation plan, or feature description and wants to break it into Jira tickets for distribution to people or agents. Also triggered by "create tickets from this plan" or "file tickets for this spec".
Use when work on a ticket has reached a stage boundary — started working, sent for review, completed, blocked, or reopened. Handles status transition, comment, and field updates in one pass.
Use when the user mentions Jira, issues, tickets, sprints, work items, or any project management task involving Atlassian. Routes to workflow skills or delegates to jira-agent subagent.
Use when the user assigns a ticket to the agent, says "work on this ticket", "pick up this item", or wants the agent to claim a ticket and execute against it. Procedural harness that wraps any work with Jira discipline.
Use when a project's workflow statuses need to be configured for the first time, or refreshed. Triggered automatically by jira-progress when a project is missing from config/workflows.json, or manually via "configure workflow", "set up statuses for X", "refresh workflow for X".
This skill should be used when the user asks to "refresh Todoist", "sync Todoist", "list my projects", "show my labels", "update Todoist structure", "show my Todoist sections", or when a todoist-agent operation fails with a not-found error for a project, label, or section name. Fetches and caches Todoist structure for name-to-ID resolution.
| name | jira-context |
| description | Use when needing to understand a Jira ticket before acting on it — pulling context for an agent, reviewing a ticket's full picture, loading ticket details for decision-making, or when asked to "show me" or "what's in" a ticket. |
Fetches a ticket and its full picture — subtasks, comments, linked context — and packages it as a structured block for agent or human consumption.
acli jira workitem view KEY-123 --fields "key,summary,status,assignee,priority,issuetype,description" --json
acli jira workitem search --jql "parent = KEY-123" --fields "key,summary,status" --csv
acli jira workitem comment-list --key "KEY-123"
## KEY-123: "Title"
Status: In Progress | Priority: High | Assignee: user@email.com
Type: Story | Sprint: Sprint 42
### Description
[full description text]
### Subtasks
- KEY-124: "Subtask A" [Done]
- KEY-125: "Subtask B" [In Progress]
### Recent Comments (last 5)
- Alice (2d ago): "Blocked on API access"
- Bob (1d ago): "Unblocked, credentials shared"
Always use this output format. Downstream skills depend on its structure.
If the ticket has no subtasks or comments, omit those sections rather than showing empty headings.
If ACLI returns exit code != 0:
acli jira auth login --web --site <site>.atlassian.netbrew tap atlassian/homebrew-acli && brew install acli