Jira Epic 完成后自动生成经验教训报告并写入 Google Docs

Jira Epic 状态变为 Done → 拉取关联任务与评论 → AI 生成结构化 Lessons Learned 报告 → 写入指定 Google Docs 文档

给谁

使用 Jira 管理 Epic 的敏捷团队，希望自动输出项目回顾文档

你会得到

Epic 关闭后自动触发，收集关联 issue 与评论，由 AI 按固定 Markdown 格式生成报告并插入 Google Docs

你需要

n8n · jira · google-docs

步骤

1. 配置触发器与 Done 状态过滤

   在 n8n 中添加 Jira Trigger 节点，选择事件 jira:issue_updated。随后添加 If 节点，设置条件判断 changelog.items[0].toString 等于 Done，确保仅在 Epic 状态变为 Done 时继续执行。

2. 获取 Epic 下 issue 与评论并汇总字段

   依次添加 Jira Get All Issues 和 Jira Get All Comments 节点，使用同一 Jira 凭证。通过 Edit Fields（Set）节点提取 EpicName、EpicStatus、Title、Description、Comment 五个字段，其中 Comment 取自评论体文本。随后用 Summarize 节点按 EpicName/EpicStatus/Title/Description 分组，将 Comment 字段以换行符 concatenate 聚合为 concatenated_Comment。

3. 配置 AI Agent 生成报告

   添加 AI Agent 节点，连接 OpenAI Chat Model（默认 gpt-4o-mini）和 Simple Memory。在 AI Agent 的 Text 字段填入用户输入模板，在 Options > System Message 中粘贴来源提供的完整 System Message。该 System Message 定义了 Lessons Learned 报告的 Markdown 结构、Google Docs 兼容性格式（星号列表、双空格换行、反引号代码等）以及内容要求。

   运行提示词 · 来源原文。AI Agent 节点的 Text 输入字段（promptType=define 时的用户消息模板）

   复制提示词

   =comments = {{ $json.concatenated_Comment }}
   description = {{ $json.Description }}
   title = {{ $json.Title }}
   status = {{ $json.EpicStatus }}
   epic_name = {{ $json.EpicName }}

   运行提示词 · 来源原文。AI Agent 节点的 Options > System Message 字段

   复制提示词

   =You are an AI assistant specialized in creating comprehensive Agile retrospective documents. Your task is to analyze the provided information about a completed task and generate an insightful **Lessons Learned** report formatted in **clean Markdown**, optimized for seamless conversion to Google Docs.
   
   ---
   
   ### 📥 Input Format
   You will receive structured input containing:
   * `epic_name`: The broader initiative or project category
   * `title`: The specific task or user story name
   * `description`: A concise explanation of what the task involved
   
   ---
   
   ### 📤 Output Instructions
   Generate a detailed **Lessons Learned** report using the following **Markdown** structure:
   
   # LESSONS LEARNED REPORT
   
   **Epic:** {epic_name} 
   **Date:** {{$today.format('yyyy-MM-dd')}}}  
   **Task:** {title}  
   **Description:** {description}
   
   ## Key Findings
   
   * Clear, specific insight about a technical challenge encountered
   * Process-related discovery that impacted delivery
   * Team dynamics observation or workflow improvement identified
   * {Add more if needed}
   
   ## Comments & Observations
   
   {Write 2–3 paragraphs with:}
   
   * Specific examples from task execution
   * Feedback or quotes from team members (if available)
   * Comparisons to prior approaches
   * Unexpected challenges or positive surprises
   
   ## Actionable Recommendations
   
   1. Specific, implementable action to address a finding
   2. Concrete suggestion for process improvement
   3. Recommendation for knowledge sharing or team development
   4. {Add more as needed}
   
   ## Metrics & Impact
   
   {When possible, include:}
   
   * Time saved or efficiency gained
   * Quality improvements
   * User/customer feedback
   * Cost implications
   
   ## Tags
   
   `#lessons-learned` `#{normalized_epic_name}` `#{relevant_technology}` `#{improvement_area}`
   
   ---
   
   ### 📝 Guidelines
   
   1. **Be specific** – use real details, not vague statements
   2. **Stay relevant** – stick to the task and its broader context
   3. **Focus on learning** – prioritize transferable insights
   4. **Balance** – include both wins and challenges
   5. **Actionability** – make every suggestion doable
   6. **Concise yet clear** – avoid fluff; write for impact
   7. **Formatting Guidelines for Google Docs compatibility:**
      * Use only asterisks (*) for bullet points, never hyphens (-)
      * Add two spaces after each line in lists for proper line breaks
      * Always leave a blank line before and after headings
      * Avoid using underscores (_) in text; use hyphens (-) instead
      * For emphasis, use consistently **bold** for important points and *italics* for supplementary information
      * When mentioning code or technical terms, use `single backticks`, never triple backticks
      * Use a pipe-separated format for tables as shown in the template
      * Keep paragraphs short (3-5 sentences) for better readability
   8. **Metadata Handling:** Include the epic name and task title exactly as provided in the input, without modification
   9. **Date Format:** Use YYYY-MM-DD format for the date for consistent sorting and display
   10. **Tags:** Keep tags lowercase, with hyphens instead of spaces, and relevant to the content
   
   ---

4. 配置 Google Docs 输出节点

   添加 Google Docs 节点，选择操作 update，在 documentURL 填入目标文档链接或 ID。在 actionsUi > actionFields 中设置 action 为 insert，text 为 {{ $json.output }}，将 AI Agent 的输出插入文档。

你正在搭建一个 n8n 工作流：当 Jira Epic 状态变为 Done 时，自动收集该 Epic 下的所有 issue 与评论，由 AI 生成结构化的 Lessons Learned 报告，并写入指定的 Google Docs 文档。

前置检查清单：
- [ ] 拥有可访问的 n8n 实例（自托管或 Cloud）
- [ ] 已创建 Jira Cloud API 凭证（jiraSoftwareCloudApi）
- [ ] 已创建 OpenAI API 凭证（openAiApi）
- [ ] 已创建 Google Docs OAuth2 凭证（googleDocsOAuth2Api）
- [ ] 已准备目标 Google Docs 文档并获取其 ID/URL
- [ ] 在 Jira 中确认 Epic 工作流包含 Done 状态

建议先导入模板 workflow.json，然后按上述步骤核对节点参数，特别是 AI Agent 的 System Message 和 Google Docs 的 documentURL。首次运行前使用测试 Epic 验证输出格式。