AI 每日会议更新生成器 - 自动整合 GitHub/Jira/Claude Code 生成站会报告

daily-meeting-update by davila7/claude-code-templates

127 周安装量

23,400 GitHub Stars

GitHub

安装命令

npx skills add https://github.com/davila7/claude-code-templates --skill daily-meeting-update

团队协作项目管理自动化

🇨🇳中文介绍

每日会议更新

通过交互式访谈生成每日站会/会议更新。切勿假设工具已配置——先询问。

工作流程

START
  │
  ▼
┌─────────────────────────────────────────────────────┐
│ 阶段 1: 检测并提供集成选项                          │
│ • 检查: Claude Code 历史？gh CLI？jira CLI？        │
│ • Claude Code → 拉取昨天的会话摘要                  │
│   → 用户通过多选选择相关项                          │
│ • GitHub/Jira → 询问用户，若批准则拉取              │
│ • 立即拉取数据（在访谈之前）                        │
├─────────────────────────────────────────────────────┤
│ 阶段 2: 访谈（结合洞察）                            │
│ • 将拉取的数据作为上下文展示                        │
│ • 昨天："我看到你合并了 PR #123，还有其他吗？"      │
│ • 今天：你今天要做什么？                            │
│ • 阻碍：有什么阻碍你吗？                            │
│ • 议题：会议结束时有什么要讨论的吗？                │
├─────────────────────────────────────────────────────┤
│ 阶段 3: 生成更新                                    │
│ • 结合访谈答案 + 工具数据                           │
│ • 格式化为简洁的 Markdown                           │
│ • 呈现给用户                                        │
└─────────────────────────────────────────────────────┘

阶段 1: 检测并提供集成选项

步骤 1: 静默检测

静默检查可用的集成（抑制错误，不向用户显示）：

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

联系我们

集成	检测方法
Claude Code 历史	`~/.claude/projects` 目录存在且包含 `.jsonl` 文件
GitHub CLI	`gh auth status` 执行成功
Jira CLI	`jira` 命令存在
Atlassian MCP	`mcp__atlassian__*` 工具可用
Git	在 git 仓库内

步骤 2: 提供 GitHub/Jira 集成选项（如果可用）

Claude Code 用户： 在此阶段的所有问题都使用 AskUserQuestionTool 工具。

如果 HAS_GH 或 HAS_GIT：

"我检测到你已配置 GitHub/Git。需要我拉取你最近的活动（提交、PR、代码审查）吗？"

选项：
- "是的，拉取信息"
- "不，我会手动提供所有信息"

"我应该检查哪些仓库/项目？"

选项：
- "仅当前目录"（如果在 git 仓库内）
- "我会列出仓库" → 用户提供列表

如果 HAS_JIRA_CLI 或 HAS_ATLASSIAN_MCP：

"我检测到你已配置 Jira。需要我拉取你的工单吗？"

选项：
- "是的，拉取我的工单"
- "不，我会手动提供所有信息"

步骤 3: 拉取 GitHub/Jira 数据（如果批准）

GitHub/Git — 对于每个批准的仓库：

自昨天以来用户进行的提交
用户打开/合并的 PR
用户完成的代码审查

Jira — 分配给用户且在最近 24 小时内更新的工单

关键洞察：存储结果，以便在阶段 2 的访谈中用作上下文。

步骤 4: 提供 Claude Code 历史

此集成捕获了你在 Claude Code 中处理的所有工作——对于回忆不在 git 或 Jira 中的工作很有用。

ls ~/.claude/projects/*/*.jsonl 2>/dev/null | head -1

如果 Claude Code 历史存在，询问：

"我也可以拉取你昨天的 Claude Code 会话历史。这有助于回忆不在 git/Jira 中的工作（研究、调试、规划）。需要我检查吗？"

选项：
- "是的，拉取我的 Claude Code 会话"
- "不，我已经有了所有需要的信息"

如果选择是，运行摘要脚本：

python3 ~/.claude/skills/daily-meeting-update/scripts/claude_digest.py --format json

然后通过多选展示会话：

使用 AskUserQuestionTool 并设置 multiSelect: true 让用户选择相关项：

"这是你昨天的 Claude Code 会话。选择与你的站会相关的会话："

选项（多选）：
- "修复认证错误 (backend-api)"
- "实现 OAuth 流程 (backend-api)"
- "更新主页样式 (frontend-app)"
- "研究支付提供商 (docs)"

关键洞察： 用户选择哪些会话与工作相关。个人项目或实验可以被排除。

在以下情况下不要运行摘要脚本：

用户明确说"不"拉取 Claude Code 历史
用户说他们会手动提供所有信息
~/.claude/projects 目录不存在

如果摘要脚本失败：

备用方案：静默跳过 Claude Code 集成，继续进行访谈
常见问题：未安装 Python、昨天没有会话、权限错误
不要阻塞站会流程——脚本是补充性的，不是必需的

阶段 2: 访谈（结合洞察）

Claude Code 用户： 使用 AskUserQuestionTool 工具进行访谈。这能通过结构化选项提供更好的用户体验。

使用拉取的数据作为上下文，使问题更智能。

如果拉取了数据，先展示它：

"这是我从你的活动中发现的：
- 合并了 PR #123：修复登录超时
- 在 backend-api 中有 3 次提交
- 审查了 PR #456（已批准）

你昨天还做了其他什么我遗漏的工作吗？"

如果未拉取数据：

"你昨天/自上次站会以来做了什么工作？"

如果用户回答模糊，请追问：

"你能提供更多关于 X 的细节吗？"
"你完成了什么具体的事情吗？"

"你今天要做什么工作？"

选项：
- [文本输入 - 用户自由输入]

如果拉取了 Jira 数据，你可以建议：

"我看到你分配了这些工单：
- PROJ-123: 实现 OAuth 流程（进行中）
- PROJ-456: 修复支付错误（待办）

你今天会处理这些中的任何一个吗？"

"你有什么阻碍或障碍吗？"

选项：
- "没有阻碍"
- "是的，我有阻碍" → 追问细节

问题 4: 讨论议题

"有什么议题你想在每日站会结束时提出讨论吗？"

选项：
- "不，没什么要讨论的"
- "是的" → 追问细节

议题示例：
- 需要输入的技术决策
- 与其他团队的对齐
- 关于优先级的问题
- 给团队的公告或信息

阶段 3: 生成更新

将所有信息合并成简洁的 Markdown：

# 每日更新 - [日期]

## 昨天
- [来自访谈的项]
- [如果从 GitHub/Jira 拉取，则包含相关项]

## 今天
- [来自访谈的项]

## 阻碍
- [阻碍或"无阻碍"]

## PR 与代码审查（如果从 GitHub 拉取）
- [打开的 PR]
- [合并的 PR]
- [完成的代码审查]

## Jira（如果从 Jira 拉取）
- [更新的工单]

## 讨论议题
- [议题或"无"]

---
*链接：*
- [PR 链接]
- [工单链接]

访谈是主要的 —— 工具是补充，不能替代人的上下文
访问前需征得同意 —— 从任何集成拉取数据前都要询问
上下文感知的问题 —— 在访谈期间展示拉取的数据以触发记忆（"我看到你合并了 PR #123..."）

阶段	操作	工具
1. 检测与提供	检查 gh/jira/claude 历史，询问用户，拉取数据	Bash（静默），AskUserQuestionTool*
2. 访谈	询问 4 个结合洞察的问题	AskUserQuestionTool*
3. 生成	格式化 Markdown	输出文本

*仅限 Claude Code：使用 AskUserQuestionTool 工具进行结构化提问。

Claude Code 摘要脚本

# 获取昨天的会话为 JSON
python3 ~/.claude/skills/daily-meeting-update/scripts/claude_digest.py --format json

# 获取今天的会话
python3 ~/.claude/skills/daily-meeting-update/scripts/claude_digest.py --date today --format json

# 筛选特定项目
python3 ~/.claude/skills/daily-meeting-update/scripts/claude_digest.py --project ~/my-app --format json

Claude: "让我们准备你的每日站会！我检测到已配置 GitHub。需要我拉取你最近的活动吗？"
User: "是的"

Claude: "我应该检查哪些仓库？"
User: "backend-api 和 frontend-app"

[Claude 静默拉取 GitHub 数据]

Claude: "我也可以拉取你昨天的 Claude Code 会话历史。需要我检查吗？"
User: "是的"

[Claude 运行摘要脚本，获取 5 个会话]

Claude: "这是你昨天的 Claude Code 会话。选择与你的站会相关的会话："
  ☑ 修复认证超时错误 (backend-api)
  ☑ 实现 OAuth 流程 (backend-api)
  ☐ 更新个人博客 (blog) - [用户取消勾选个人项目]
  ☑ 研究支付提供商 (docs)
  ☐ 调试测试失败 (experiments)

User: [选择 3 个相关项]

Claude: "这是我从你的活动中发现的：
- 合并了 PR #120：修复登录超时
- 打开了 PR #125：feat: 添加 OAuth 流程
- 审查了 PR #123（已批准），PR #456（请求更改）
- 在 backend-api 中有 5 次提交
- [来自 Claude Code] 修复认证超时错误
- [来自 Claude Code] 实现 OAuth 流程
- [来自 Claude Code] 研究支付提供商

你昨天还做了其他什么我遗漏的工作吗？"
User: "不，就这些了"

Claude: "你今天要做什么工作？"
User: "继续 OAuth 功能并部署到 staging 环境"

Claude: "有什么阻碍吗？"
User: "没有"

Claude: "有什么议题想在站会结束时提出讨论吗？"
User: "是的，我想讨论新支付模块的架构"

[Claude 生成更新]

# 每日更新 - 2026-01-22

## 昨天
- 处理了认证功能
- 研究了支付提供商
- 合并了 PR #120 (fix: 登录超时)
- 打开了 PR #125 (feat: 添加 OAuth 流程)

## 今天
- 继续 OAuth 功能
- 部署到 staging 环境

## 阻碍
- 无阻碍

## PR 与代码审查
- **打开：** PR #125 - feat: 添加 OAuth 流程
- **合并：** PR #120 - fix: 登录超时
- **审查：** PR #123（已批准），PR #456（请求更改）

## 讨论议题
- 新支付模块的架构

---
*链接：*
- https://github.com/org/repo/pull/125
- https://github.com/org/repo/pull/120

避免	原因（专家知识）	替代做法
未经询问就运行 gh/jira	用户可能有可见的个人仓库，或者处于他们不想暴露的敏感项目上下文中	始终先询问，让用户选择仓库
假设当前目录是唯一项目	开发者通常同时处理 2-5 个仓库（前端、后端、基础设施）	询问"你在处理哪些项目？"
即使有工具数据也跳过访谈	工具捕获了发生了什么，但遗漏了原因和上下文（研究、会议、规划）	访谈是主要的，工具是补充的
在 4 个问题都问完前生成更新	用户可能有关键的阻碍或讨论议题，这会改变叙述	完成访谈，然后生成
包含原始的提交信息	提交信息通常很隐晦（"fix"、"wip"），无法讲述完整故事	总结成人类可读的结果
在访谈后请求数据	在访谈期间展示洞察能使问题更智能（"我看到你合并了 PR #123，还有其他吗？"）	先拉取数据，然后在有上下文的情况下进行访谈

绝对不要假设工具已配置 —— 许多开发者安装了 gh 但未认证，或者 jira CLI 指向错误的实例
绝对不要跳过"讨论议题"问题 —— 这通常是站会最有价值的部分，而工具无法捕获
绝对不要生成超过 15 个要点 —— 站会应该能在 2 分钟内读完；冗长的更新会失去听众
绝对不要包含没有上下文的工单/PR 编号 —— "PROJ-123" 毫无意义；始终包含标题或摘要
绝对不要从用户未明确批准的仓库拉取数据 —— 即使你能看到其他仓库，也要尊重边界

🇺🇸English

Daily Meeting Update

Generate a daily standup/meeting update through an interactive interview. Never assume tools are configured—ask first.

Workflow

START
  │
  ▼
┌─────────────────────────────────────────────────────┐
│ Phase 1: DETECT & OFFER INTEGRATIONS                │
│ • Check: Claude Code history? gh CLI? jira CLI?     │
│ • Claude Code → Pull yesterday's session digest     │
│   → User selects relevant items via multiSelect     │
│ • GitHub/Jira → Ask user, pull if approved          │
│ • Pull data NOW (before interview)                  │
├─────────────────────────────────────────────────────┤
│ Phase 2: INTERVIEW (with insights)                  │
│ • Show pulled data as context                       │
│ • Yesterday: "I see you merged PR #123, what else?" │
│ • Today: What will you work on?                     │
│ • Blockers: Anything blocking you?                  │
│ • Topics: Anything to discuss at end of meeting?    │
├─────────────────────────────────────────────────────┤
│ Phase 3: GENERATE UPDATE                            │
│ • Combine interview answers + tool data             │
│ • Format as clean Markdown                          │
│ • Present to user                                   │
└─────────────────────────────────────────────────────┘

Phase 1: Detect & Offer Integrations

Step 1: Silent Detection

Check for available integrations silently (suppress errors, don't show to user):

Integration	Detection
Claude Code History	`~/.claude/projects` directory exists with `.jsonl` files
GitHub CLI	`gh auth status` succeeds
Jira CLI	`jira` command exists
Atlassian MCP	`mcp__atlassian__*` tools available
Git	Inside a git repository

Step 2: Offer GitHub/Jira Integrations (if available)

Claude Code users: Use AskUserQuestionTool tool for all questions in this phase.

GitHub/Git:

If HAS_GH or HAS_GIT:

"I detected you have GitHub/Git configured. Want me to pull your recent activity (commits, PRs, reviews)?"

Options:
- "Yes, pull the info"
- "No, I'll provide everything manually"

If yes:

"Which repositories/projects should I check?"

Options:
- "Just the current directory" (if in a git repo)
- "I'll list the repos" → user provides list

Jira:

If HAS_JIRA_CLI or HAS_ATLASSIAN_MCP:

"I detected you have Jira configured. Want me to pull your tickets?"

Options:
- "Yes, pull my tickets"
- "No, I'll provide everything manually"

Step 3: Pull GitHub/Jira Data (if approved)

GitHub/Git — For each approved repo:

Commits by user since yesterday
PRs opened/merged by user
Reviews done by user

Jira — Tickets assigned to user, updated in last 24h

Key insight : Store results to use as context in Phase 2 interview.

Step 4: Offer Claude Code History

This integration captures everything you worked on with Claude Code — useful for recalling work that isn't in git or Jira.

Detection:

ls ~/.claude/projects/*/*.jsonl 2>/dev/null | head -1

If Claude Code history exists, ask:

"I can also pull your Claude Code session history from yesterday. This can help recall work that isn't in git/Jira (research, debugging, planning). Want me to check?"

Options:
- "Yes, pull my Claude Code sessions"
- "No, I have everything I need"

If yes, run the digest script:

python3 ~/.claude/skills/daily-meeting-update/scripts/claude_digest.py --format json

Then present sessions with multiSelect:

Use AskUserQuestionTool with multiSelect: true to let user pick relevant items:

"Here are your Claude Code sessions from yesterday. Select the ones relevant to your standup:"

Options (multiSelect):
- "Fix authentication bug (backend-api)"
- "Implement OAuth flow (backend-api)"
- "Update homepage styles (frontend-app)"
- "Research payment providers (docs)"

Key insight: User selects which sessions are work-related. Personal projects or experiments can be excluded.

Do NOT run digest script when:

User explicitly says "No" to Claude Code history
User says they'll provide everything manually
~/.claude/projects directory doesn't exist

If digest script fails:

Fallback: Skip Claude Code integration silently, proceed with interview
Common issues: Python not installed, no sessions from yesterday, permission errors
Do NOT block the standup flow — the script is supplemental, not required

Phase 2: Interview (with insights)

Claude Code users: Use AskUserQuestionTool tool to conduct the interview. This provides a better UX with structured options.

Use pulled data as context to make questions smarter.

Question 1: Yesterday

If data was pulled , show it first:

"Here's what I found from your activity:
- Merged PR #123: fix login timeout
- 3 commits in backend-api
- Reviewed PR #456 (approved)

Anything else you worked on yesterday that I missed?"

If no data pulled:

"What did you work on yesterday/since the last standup?"

If user response is vague, ask follow-up:

"Can you give more details about X?"
"Did you complete anything specific?"

Question 2: Today

"What will you work on today?"

Options:
- [Text input - user types freely]

If Jira data was pulled , you can suggest:

"I see you have these tickets assigned:
- PROJ-123: Implement OAuth flow (In Progress)
- PROJ-456: Fix payment bug (To Do)

Will you work on any of these today?"

Question 3: Blockers

"Do you have any blockers or impediments?"

Options:
- "No blockers"
- "Yes, I have blockers" → follow-up for details

Question 4: Topics for Discussion

"Any topic you want to bring up at the end of the daily?"

Options:
- "No, nothing to discuss"
- "Yes" → follow-up for details

Examples of topics:
- Technical decision that needs input
- Alignment with another team
- Question about prioritization
- Announcement or info for the team

Phase 3: Generate Update

Combine all information into clean Markdown:

# Daily Update - [DATE]

## Yesterday
- [Items from interview]
- [Items from GitHub/Jira if pulled]

## Today
- [Items from interview]

## Blockers
- [Blockers or "No blockers"]

## PRs & Reviews (if pulled from GitHub)
- [PRs opened]
- [PRs merged]
- [Reviews done]

## Jira (if pulled from Jira)
- [Tickets updated]

## Topics for Discussion
- [Topics or "None"]

---
*Links:*
- [PR links]
- [Ticket links]

Core Principles

Interview is primary — Tools supplement, they don't replace human context
Consent before access — Always ask before pulling from any integration
Context-aware questions — Show pulled data during interview to trigger memory ("I see you merged PR #123...")

Quick Reference

Phase	Action	Tool
1. Detect & Offer	Check gh/jira/claude history, ask user, pull data	Bash (silent), AskUserQuestionTool*
2. Interview	Ask 4 questions with insights	AskUserQuestionTool*
3. Generate	Format Markdown	Output text

*Claude Code only: Use AskUserQuestionTool tool for structured questions.

Claude Code Digest Script

# Get yesterday's sessions as JSON
python3 ~/.claude/skills/daily-meeting-update/scripts/claude_digest.py --format json

# Get today's sessions
python3 ~/.claude/skills/daily-meeting-update/scripts/claude_digest.py --date today --format json

# Filter to specific project
python3 ~/.claude/skills/daily-meeting-update/scripts/claude_digest.py --project ~/my-app --format json

Example Session

Claude: "Let's prepare your daily! I detected GitHub configured. Want me to pull your recent activity?"
User: "Yes"

Claude: "Which repos should I check?"
User: "backend-api and frontend-app"

[Claude pulls GitHub data silently]

Claude: "I can also pull your Claude Code session history from yesterday. Want me to check?"
User: "Yes"

[Claude runs digest script, gets 5 sessions]

Claude: "Here are your Claude Code sessions from yesterday. Select the ones relevant to your standup:"
  ☑ Fix authentication timeout bug (backend-api)
  ☑ Implement OAuth flow (backend-api)
  ☐ Update personal blog (blog) - [user unchecks personal project]
  ☑ Research payment providers (docs)
  ☐ Debugging test failures (experiments)

User: [selects 3 relevant items]

Claude: "Here's what I found from your activity:
- Merged PR #120: fix login timeout
- Opened PR #125: feat: add OAuth flow
- Reviewed PR #123 (approved), PR #456 (changes requested)
- 5 commits in backend-api
- [From Claude Code] Fix authentication timeout bug
- [From Claude Code] Implement OAuth flow
- [From Claude Code] Research payment providers

Anything else you worked on yesterday that I missed?"
User: "No, that covers it"

Claude: "What will you work on today?"
User: "Continue the OAuth feature and deploy to staging"

Claude: "Any blockers?"
User: "No"

Claude: "Any topic to bring up at the end of the daily?"
User: "Yes, I want to discuss the architecture of the new payments module"

[Claude generates update]

Output Example

# Daily Update - 2026-01-22

## Yesterday
- Worked on authentication feature
- Research on payment providers
- Merged PR #120 (fix: login timeout)
- Opened PR #125 (feat: add OAuth flow)

## Today
- Continue OAuth feature
- Deploy to staging

## Blockers
- No blockers

## PRs & Reviews
- **Opened:** PR #125 - feat: add OAuth flow
- **Merged:** PR #120 - fix: login timeout
- **Reviews:** PR #123 (approved), PR #456 (changes requested)

## Topics for Discussion
- Architecture of the new payments module

---
*Links:*
- https://github.com/org/repo/pull/125
- https://github.com/org/repo/pull/120

Anti-Patterns

Avoid	Why (Expert Knowledge)	Instead
Run gh/jira without asking	Users may have personal repos visible, or be in a sensitive project context they don't want exposed	Always ask first, let user choose repos
Assume current directory is the only project	Developers often work on 2-5 repos simultaneously (frontend, backend, infra)	Ask "Which projects are you working on?"
Skip interview even with tool data	Tools capture WHAT happened but miss WHY and context (research, meetings, planning)	Interview is primary, tools supplement
Generate update before all 4 questions	User might have critical blocker or discussion topic that changes the narrative	Complete interview, then generate
Include raw commit messages	Commit messages are often cryptic ("fix", "wip") and don't tell the story	Summarize into human-readable outcomes
Ask for data after interview	Showing insights during interview makes questions smarter ("I see you merged PR #123, anything else?")	Pull data first, then interview with context

NEVER

NEVER assume tools are configured — Many devs have gh installed but not authenticated, or jira CLI pointing to wrong instance
NEVER skip the "Topics for Discussion" question — This is often the most valuable part of standup that tools can't capture
NEVER generate more than 15 bullets — Standup should be <2 minutes to read; long updates lose the audience
NEVER include ticket/PR numbers without context — "PROJ-123" means nothing; always include title or summary
NEVER pull data from repos user didn't explicitly approve — Even if you can see other repos, respect boundaries

Weekly Installs

106

Repository

davila7/claude-…emplates

GitHub Stars

22.6K

First Seen

Jan 25, 2026

Security Audits

Gen Agent Trust HubWarn SocketPass SnykFail

Installed on

claude-code86

opencode85

cursor83

gemini-cli77

antigravity72

codex72

通过 LiteLLM 代理让 Claude Code 对接 GitHub Copilot 运行 | 高级变通方案指南

36,300 周安装