重要前提
安装AI Skills的关键前提是:必须科学上网,且开启TUN模式,这一点至关重要,直接决定安装能否顺利完成,在此郑重提醒三遍:科学上网,科学上网,科学上网。查看完整安装教程 →
jira-epic by gentleman-programming/gentleman-skills
npx skills add https://github.com/gentleman-programming/gentleman-skills --skill jira-epic在创建以下类型的 Jira 史诗时使用此技能:
# {史诗标题}
**Figma:** {如果有 Figma 链接则提供}
## 功能概述
{2-3 段描述,说明此功能的作用以及为何需要它}
## 需求
### {第 1 部分:主要功能领域}
#### {子部分}
- 需求 1
- 需求 2
- 需求 3
#### {另一个子部分}
- 需求 1
- 需求 2
### {第 2 部分:另一个主要领域}
#### {子部分}
- 需求 1
- 需求 2
## 技术考量
### 性能
- {性能要求 1}
- {性能要求 2}
### 数据集成
- {数据源}
- {集成点}
### UI 组件
- {组件 1}
- {组件 2}
## 实施清单
- [ ] {主要交付成果 1}
- [ ] {主要交付成果 2}
- [ ] {主要交付成果 3}
## 图表
{用于架构、流程、数据模型等的 Mermaid 图表}
格式:[EPIC] 功能名称
示例:
[EPIC] 发现视图[EPIC] 多租户支持[EPIC] 合规仪表板重新设计广告位招租
在这里展示您的产品或服务
触达数万 AI 开发者,精准高效
[EPIC] GovCloud 支持简要但完整地描述:
按功能领域组织:
始终包含:
将成为独立任务的高层级交付成果:
使用 Mermaid 绘制:
graph TB
UI[UI Components] --> API[API Endpoints]
API --> SDK[Prowler SDK]
SDK --> Cloud[Cloud Providers]
sequenceDiagram
User->>UI: Apply filters
UI->>API: GET /findings?filters
API->>DB: Query findings
DB-->>API: Results
API-->>UI: JSON response
UI-->>User: Render table
stateDiagram-v2
[*] --> Pending
Pending --> InProgress: Start triage
InProgress --> Resolved: Mark resolved
InProgress --> Pending: Reset
Resolved --> [*]
erDiagram
FINDING ||--o{ RESOURCE : affects
FINDING }|--|| CHECK : "belongs to"
RESOURCE }|--|| ACCOUNT : "belongs to"
ACCOUNT }|--|| PROVIDER : "belongs to"
创建史诗后,使用 jira-task 技能生成独立任务:
从史诗 [EPIC] 发现视图,创建:
[FEATURE] 带分页的发现表格 (UI)[FEATURE] 发现过滤器 - 提供商和账户 (UI)[FEATURE] 发现详情面板 - 概览标签页 (UI)[FEATURE] 发现详情面板 - 资源标签页 (UI)[FEATURE] 发现批量操作 - 静音/抑制 (API + UI)[FEATURE] 发现搜索功能 (API + UI)始终在每个任务中指定:
## 相关任务
- Epic: [EPIC] Findings View
- Blocked by: [task if any]
- Blocks: [task if any]
当提供 Figma 链接时:
https://www.figma.com/design/xxx?node-id=1830-44712&m=dev## Epic: [EPIC] {标题}
{遵循模板的完整史诗内容}
---
## 建议任务
基于此史诗,创建以下任务:
| # | 标题 | 组件 | 被阻塞于 |
|---|-------|-----------|------------|
| 1 | [FEATURE] 任务名称 | API | - |
| 2 | [FEATURE] 任务名称 | UI | 任务 1 |
| 3 | [FEATURE] 任务名称 | UI | 任务 2 |
您希望我生成完整的任务描述吗?
[EPIC] 功能名称 格式关键: 所有输出必须是 Markdown 格式,可以直接粘贴到 Jira 中。
# 表示史诗标题,## 表示主要章节,### 表示子章节**粗体** 进行强调- [ ] 表示复选框反引号 表示文件路径、命令和代码引用--- 分隔史诗和建议任务关键: 通过 MCP 创建史诗时,请使用以下确切参数:
{
"project_key": "PROWLER",
"summary": "[EPIC] Feature name",
"issue_type": "Epic",
"additional_fields": {
"customfield_10359": {"value": "UI"}
}
}
customfield_10359(团队)字段是必需的。选项:
"UI" - 前端史诗"API" - 后端史诗"SDK" - Prowler SDK 史诗重要: 该项目使用 customfield_10363(工作项描述)字段,而不是标准的 description 字段,用于在 UI 中显示。
关键: 使用 Jira Wiki 标记,而不是 Markdown:
h2. 代替 ##*text* 表示粗体,代替 **text*** item 表示项目符号(相同)** subitem 表示嵌套项目符号创建史诗后,使用以下内容更新描述:
{
"customfield_10363": "h2. Feature Overview\n\n{overview}\n\nh2. Requirements\n\n*{Section 1}*\n* {requirement 1}\n* {requirement 2}\n\n*{Section 2}*\n* {requirement 1}\n* {requirement 2}\n\nh2. Technical Considerations\n\n*Performance:*\n* {consideration 1}\n\n*Data Integration:*\n* {consideration 2}\n\nh2. Implementation Checklist\n\n* [ ] {deliverable 1}\n* [ ] {deliverable 2}\n* [ ] {deliverable 3}"
}
创建子任务时,使用史诗键作为父级:
{
"additional_fields": {
"parent": "PROWLER-XXX"
}
}
Backlog (10037) → To Do (14) → In Progress (11) → Done (21)
→ Blocked (10)
创建史诗:
mcp__mcp-atlassian__jira_create_issue (issue_type: "Epic")
更新工作项描述:
mcp__mcp-atlassian__jira_update_issue with customfield_10363
创建子任务:
mcp__mcp-atlassian__jira_create_issue with parent: EPIC-KEY
分配和转换:
mcp__mcp-atlassian__jira_update_issue (assignee)
mcp__mcp-atlassian__jira_transition_issue (status)
jira, epic, feature, initiative, prowler, large feature
每周安装次数
45
仓库
GitHub 星标数
365
首次出现
2026年1月24日
安全审计
安装于
opencode40
codex32
gemini-cli32
cursor32
claude-code30
github-copilot30
Use this skill when creating Jira epics for:
# {Epic Title}
**Figma:** {figma link if available}
## Feature Overview
{2-3 paragraph description of what this feature does and why it's needed}
## Requirements
### {Section 1: Major Functionality Area}
#### {Subsection}
- Requirement 1
- Requirement 2
- Requirement 3
#### {Another Subsection}
- Requirement 1
- Requirement 2
### {Section 2: Another Major Area}
#### {Subsection}
- Requirement 1
- Requirement 2
## Technical Considerations
### Performance
- {Performance requirement 1}
- {Performance requirement 2}
### Data Integration
- {Data source}
- {Integration points}
### UI Components
- {Component 1}
- {Component 2}
## Implementation Checklist
- [ ] {Major deliverable 1}
- [ ] {Major deliverable 2}
- [ ] {Major deliverable 3}
## Diagrams
{Mermaid diagrams for architecture, flow, data model, etc.}
Format: [EPIC] Feature Name
Examples:
[EPIC] Findings View[EPIC] Multi-tenant Support[EPIC] Compliance Dashboard Redesign[EPIC] GovCloud SupportBrief but complete description of:
Organized by functional area:
Always include:
High-level deliverables that will become individual tasks:
Use Mermaid for:
graph TB
UI[UI Components] --> API[API Endpoints]
API --> SDK[Prowler SDK]
SDK --> Cloud[Cloud Providers]
sequenceDiagram
User->>UI: Apply filters
UI->>API: GET /findings?filters
API->>DB: Query findings
DB-->>API: Results
API-->>UI: JSON response
UI-->>User: Render table
stateDiagram-v2
[*] --> Pending
Pending --> InProgress: Start triage
InProgress --> Resolved: Mark resolved
InProgress --> Pending: Reset
Resolved --> [*]
erDiagram
FINDING ||--o{ RESOURCE : affects
FINDING }|--|| CHECK : "belongs to"
RESOURCE }|--|| ACCOUNT : "belongs to"
ACCOUNT }|--|| PROVIDER : "belongs to"
After creating the epic, generate individual tasks using the jira-task skill:
From epic [EPIC] Findings View, create:
[FEATURE] Findings table with pagination (UI)[FEATURE] Findings filters - provider and account (UI)[FEATURE] Findings detail panel - Overview tab (UI)[FEATURE] Findings detail panel - Resources tab (UI)[FEATURE] Findings bulk actions - mute/suppress (API + UI)[FEATURE] Findings search functionality (API + UI)Always specify in each task:
## Related Tasks
- Epic: [EPIC] Findings View
- Blocked by: [task if any]
- Blocks: [task if any]
When Figma links are provided:
https://www.figma.com/design/xxx?node-id=1830-44712&m=dev## Epic: [EPIC] {Title}
{Full epic content following template}
---
## Suggested Tasks
Based on this epic, create the following tasks:
| # | Title | Component | Blocked By |
|---|-------|-----------|------------|
| 1 | [FEATURE] Task name | API | - |
| 2 | [FEATURE] Task name | UI | Task 1 |
| 3 | [FEATURE] Task name | UI | Task 2 |
Would you like me to generate the full task descriptions?
[EPIC] Feature Name formatCRITICAL: All output MUST be in Markdown format, ready to paste into Jira.
# for epic title, ## for main sections, ### for subsections**bold** for emphasis- [ ] for checkboxes in Implementation Checklistbackticks for file paths, commands, and code references--- to separate epic from suggested tasksCRITICAL: When creating epics via MCP, use these exact parameters:
{
"project_key": "PROWLER",
"summary": "[EPIC] Feature name",
"issue_type": "Epic",
"additional_fields": {
"customfield_10359": {"value": "UI"}
}
}
The customfield_10359 (Team) field is REQUIRED. Options:
"UI" - Frontend epics"API" - Backend epics"SDK" - Prowler SDK epicsIMPORTANT: The project uses customfield_10363 (Work Item Description) instead of the standard description field for display in the UI.
CRITICAL: Use Jira Wiki markup , NOT Markdown:
h2. instead of ##*text* for bold instead of **text*** item for bullets (same)** subitem for nested bulletsAfter creating the epic, update the description with:
{
"customfield_10363": "h2. Feature Overview\n\n{overview}\n\nh2. Requirements\n\n*{Section 1}*\n* {requirement 1}\n* {requirement 2}\n\n*{Section 2}*\n* {requirement 1}\n* {requirement 2}\n\nh2. Technical Considerations\n\n*Performance:*\n* {consideration 1}\n\n*Data Integration:*\n* {consideration 2}\n\nh2. Implementation Checklist\n\n* [ ] {deliverable 1}\n* [ ] {deliverable 2}\n* [ ] {deliverable 3}"
}
When creating child tasks, use the epic key as parent:
{
"additional_fields": {
"parent": "PROWLER-XXX"
}
}
Backlog (10037) → To Do (14) → In Progress (11) → Done (21)
→ Blocked (10)
mcp__mcp-atlassian__jira_create_issue (issue_type: "Epic")
mcp__mcp-atlassian__jira_update_issue with customfield_10363
mcp__mcp-atlassian__jira_create_issue with parent: EPIC-KEY
mcp__mcp-atlassian__jira_update_issue (assignee)
mcp__mcp-atlassian__jira_transition_issue (status)
jira, epic, feature, initiative, prowler, large feature
Weekly Installs
45
Repository
GitHub Stars
365
First Seen
Jan 24, 2026
Security Audits
Gen Agent Trust HubPassSocketPassSnykPass
Installed on
opencode40
codex32
gemini-cli32
cursor32
claude-code30
github-copilot30
论文复现上下文解析器 - 解决AI论文复现中的关键信息缺口与冲突检测
28,400 周安装