meeting-cost-calculator

作者: microsoft

计算每周或每月的会议时间——总时长、占工作时间的百分比、参会人总时长,并找出最昂贵的定期会议。

npx skills add https://github.com/microsoft/work-iq --skill meeting-cost-calculator

💰 Meeting Cost Calculator

Analyzes your calendar to compute exactly how much time you spend in meetings over a given period. Breaks down total hours, percentage of your work week consumed, average meeting duration, attendee-hours (a proxy for organizational cost), and ranks your most expensive recurring meetings so you can make data-driven decisions about which to keep, shorten, or cancel.

When to Use

  • "How much time do I spend in meetings?"
  • "Calculate my meeting load for this week"
  • "What percentage of my time is in meetings this month?"
  • "Show me my most expensive meetings"
  • "Analyze my meeting costs for the last 4 weeks"
  • "Which recurring meetings consume the most time?"

Instructions

Step 1: Get User Profile, Timezone, and Working Hours

workiq-ask (
  question: "What is my display name, email address, time zone, and configured working hours (start time, end time, and working days of the week)?"
)

Extract displayName, mail, timeZone, working hours start/end, and working days from the response. Calculate the total available work hours per week (e.g. 5 days × 8 hours = 40 hours).

Step 2: Pull Calendar Data for the Period

Determine the analysis period from the user's request (default: current week). For multi-week analysis, pull the full range.

workiq-ask (
  question: "List all my calendar events from {period start date} to {period end date} including the subject, start time, end time, attendees list with count, whether it is an all-day event, whether it is cancelled, recurrence details, organizer, showAs status, and event type."
)

Filter out cancelled events, all-day events, and events where showAs is "free". These do not count as meeting time.

Step 3: Compute Core Metrics

For each qualifying event, compute:

  • Duration in minutes: (end - start)
  • Attendee count: length of attendees array + 1 (organizer)
  • Attendee-hours: duration_hours × attendee_count
  • Is recurring: check if type is "seriesMaster" or recurrence is populated

Aggregate across the period:

  • Total meeting hours: sum of all durations
  • Meeting count: total number of events
  • Average meeting duration: total hours / meeting count
  • Work hours in period: working days × daily work hours
  • Meeting percentage: (total meeting hours / work hours) × 100
  • Total attendee-hours: sum of all attendee-hours

Step 4: Break Down by Day

Group meetings by day and compute daily meeting load:

  • Total hours per day
  • Number of meetings per day
  • Longest consecutive meeting streak

Build a per-day bar chart for visual display.

Step 5: Identify Most Expensive Recurring Meetings

Group recurring meeting instances by their series subject. For each series:

  • Count occurrences in the period
  • Total hours consumed
  • Average attendee count
  • Total attendee-hours across all instances
  • Frequency (daily, weekly, biweekly, monthly)

Sort by total hours descending. The top entries are the "most expensive" meetings.

Step 6: Generate Cost Insights

Compute actionable insights:

  • If meeting % > 60%: flag as "⚠️ Meeting overload — consider auditing recurring meetings"
  • If any single recurring meeting > 3 hours/week: flag as a reduction candidate
  • If average meeting length > 45 min: suggest defaulting to 25-minute or 50-minute meetings
  • If any day has > 6 hours of meetings: flag that day as a risk for burnout

Output Format

💰 MEETING COST CALCULATOR
═══════════════════════════════════════════════════════

📅 Period: July 14 – 18, 2025 (1 week)
👤 User: Firstname1 Lastname1
⏰ Work week: 40 hours (Mon–Fri, 9 AM – 5 PM PST)

───────────────────────────────────────────────────────
📊 SUMMARY
───────────────────────────────────────────────────────

  Total meeting time:     22.5 hours
  Meeting count:          18 meetings
  Work time in meetings:  56.3%  ⚠️  Above 50% threshold
  Average duration:       1h 15m
  Total attendee-hours:   112.5 hours
  Longest meeting streak: 3h 30m (Tuesday)

───────────────────────────────────────────────────────
📊 DAILY MEETING LOAD
───────────────────────────────────────────────────────

  Mon │ ████████████░░░░░░░░  3.0h (37%)   4 meetings
  Tue │ ██████████████████░░  6.5h (81%)   5 meetings ⚠️
  Wed │ ██████████████░░░░░░  5.0h (62%)   4 meetings
  Thu │ ████████░░░░░░░░░░░░  4.0h (50%)   3 meetings
  Fri │ ████████░░░░░░░░░░░░  4.0h (50%)   2 meetings
      └────────────────────
        0h   2h   4h   6h   8h

───────────────────────────────────────────────────────
🔄 MOST EXPENSIVE RECURRING MEETINGS
───────────────────────────────────────────────────────

  #  Meeting                  Freq     Dur    Attend  Hours/Wk
  ─────────────────────────────────────────────────────────────
  1  Sprint Planning          Weekly   2h 00m    12   2.0h  ⚠️
  2  Team Standup             Daily    0h 30m     8   2.5h  ⚠️
  3  Design Review            Weekly   1h 30m     6   1.5h
  4  1:1 with Manager         Weekly   0h 30m     2   0.5h
  5  All-Hands                Weekly   1h 00m    50   1.0h

───────────────────────────────────────────────────────
💡 INSIGHTS & RECOMMENDATIONS
───────────────────────────────────────────────────────

  ⚠️  56% of your work week is in meetings — aim for < 40%
  ⚠️  Tuesday is dangerously overloaded (6.5h of meetings)
  💡  Team Standup costs 2.5h/wk across the team — consider
      async standups 2 days/week to save 1h
  💡  Sprint Planning at 2h may benefit from a 90-min timebox
  💡  Consider "No Meeting Tuesday" to reclaim 6.5h

Parameters

ParameterRequiredDefaultDescription
periodNo"this week"Analysis period: "this week", "last week", "this month", "last 4 weeks", or custom date range
startDateNoAutoCustom start date in ISO 8601
endDateNoAutoCustom end date in ISO 8601
includeDeclinedNofalseWhether to include meetings you declined
meetingOverloadThresholdNo50Percentage above which to flag meeting overload

Required MCP Tools

MCP ServerToolPurpose
workiq (Local WorkIQ CLI)askGet user profile, timezone, working hours, and retrieve calendar events for the analysis period

Tips

  • Run monthly to track trends — are your meetings growing or shrinking?
  • Share the output with your manager to justify meeting pruning.
  • Use the "Most Expensive" table to decide which recurring meetings to audit, shorten, or cancel.
  • Attendee-hours is the best proxy for organizational cost — a 1-hour meeting with 10 people costs 10 person-hours.
  • After identifying days with low meeting load, consider blocking focus time on your calendar.
  • Events with showAs: "free" are excluded since they don't block your calendar.

Examples

Example 1: Analyze the current week

"How much of my time is in meetings this week?"

Claude fetches your calendar for the current Mon–Fri, computes total meeting hours against your configured work hours, and returns a full breakdown including daily load chart, recurring-meeting rankings, and any overload flags.


Example 2: Monthly analysis with custom threshold

"Analyze my meeting costs for this month and flag anything above 40% of my time."

Claude sets meetingOverloadThreshold to 40, pulls the full month of calendar data, aggregates all metrics, and highlights any day or week where meetings exceeded that threshold — plus surfaces the top recurring meetings driving the highest attendee-hours.


Example 3: Last 4 weeks trend

"Which recurring meetings have cost me the most time over the last 4 weeks?"

Claude retrieves four weeks of calendar data, groups recurring series across the entire range, and ranks them by total hours consumed. Useful for preparing a meeting-audit conversation with your manager or team.

Error Handling

Calendar data unavailable or empty

If ask returns no events for the requested period, confirm the date range is correct and that the authenticated account has calendar read permissions. Remind the user that events on secondary or shared calendars are not included unless those calendars are surfaced in the primary view.

Timezone or working-hours settings missing

If ask returns no working hours information, fall back to a standard 40-hour work week (Mon–Fri, 9 AM – 5 PM UTC) and inform the user that results may not reflect their actual schedule. Prompt them to verify timezone settings in their Microsoft 365 profile.

Attendee count is zero or undefined

Some events (e.g., personal blocks or private meetings) may omit the attendees field. In these cases, default attendee count to 1 (the user only) so attendee-hours still accumulate correctly. Flag these events in a footnote as "attendee data unavailable."

Period spans a holiday or non-working period

If work hours for a given day are zero (weekend, public holiday), exclude that day from the work-hours denominator to avoid inflating the meeting percentage. If the entire requested period falls outside normal working days, notify the user and suggest choosing a different range.

Insufficient permissions

If MCP tool calls return a 403 or permission error, instruct the user to ensure the Claude integration has Calendars.Read scope granted in their Microsoft 365 tenant admin settings.

来自 microsoft 的更多技能

oss-growth
microsoft
OSS增长黑客角色
agent-framework-azure-ai-py
microsoft
使用Microsoft Agent Framework Python SDK(agent-framework-azure-ai)构建Azure AI Foundry代理。在创建使用AzureAIAgentsProvider的持久化代理、使用托管工具(代码解释器、文件搜索、网络搜索)、集成MCP服务器、管理对话线程或实现流式响应时使用。涵盖函数工具、结构化输出和多工具代理。
development
airunway-aks-setup
microsoft
在AKS上设置AI Runway——从裸集群到运行模型。涵盖集群验证、控制器安装、GPU评估、提供商设置和首次部署。适用场景:“设置AI Runway”、“接入AKS集群”、“安装AI Runway”、“airunway设置”、“将模型部署到AKS”、“在AKS上进行GPU推理”、“在AKS上配置KAITO”、“在AKS上运行LLM”、“在AKS上使用vLLM”、“在AKS上设置模型服务”、“AI Runway控制器”。
devops
appinsights-instrumentation
microsoft
使用Azure Application Insights对Web应用进行插桩的指南。提供遥测模式、SDK设置和配置参考。适用场景:如何对应用进行插桩、App Insights SDK、遥测模式、什么是App Insights、Application Insights指南、插桩示例、APM最佳实践。
devops
applicationinsights-web-ts
microsoft
使用Application Insights JavaScript SDK(@microsoft/applicationinsights-web)为浏览器/Web应用添加检测。用于真实用户监控(RUM)——页面视图、点击、AJAX/fetch依赖项、异常、自定义事件,以及与后端OpenTelemetry追踪关联的浏览器端GenAI代理追踪。涵盖SDK加载器脚本和npm设置、框架扩展(React、React Native、Angular)、点击分析、遥测初始化器,以及从浏览器发出的代理/工具/模型跨度所遵循的OTel GenAI语义约定。
devops
azure-ai-anomalydetector-java
microsoft
使用适用于 Java 的 Azure AI 异常检测器 SDK 构建异常检测应用程序。在实现单变量/多变量异常检测、时间序列分析或 AI 驱动的监控时使用。
development
azure-ai-language-conversations-py
microsoft
使用azure-ai-language-conversations Python SDK实现对话语言理解(CLU)。当使用ConversationAnalysisClient分析对话意图和实体、构建NLP功能或将语言理解集成到应用程序中时使用。
development
azure-ai-ml-py
microsoft
Azure Machine Learning SDK v2 for Python。用于机器学习工作区、作业、模型、数据集、计算资源和管道。 触发词:“azure-ai-ml”、“MLClient”、“工作区”、“模型注册表”、“训练作业”、“数据集”。
development