本文へ移動
cccskills
無料GitHub で公開

agile-sprint-planning

Plan and execute effective sprints using Agile methodologies. Define sprint goals, estimate user stories, manage sprint backlog, and facilitate daily standups to maximize team productivity and deliver value incrementally.

インストール方法を見る

含まれるファイル(1)

  • SKILL.md8.1 KB

SKILL.md(原文)

インストールする前に、エージェントに与えられる指示の中身を確認できます。

Agile Sprint Planning

Overview

Agile sprint planning provides a structured approach to organize work into time-boxed iterations, enabling teams to deliver value incrementally while maintaining flexibility and responding to change.

When to Use

  • Starting a new sprint cycle
  • Defining sprint goals and objectives
  • Estimating user stories and tasks
  • Managing sprint backlog prioritization
  • Handling mid-sprint changes or scope adjustments
  • Preparing sprint reviews and retrospectives
  • Training team members on Agile practices

Instructions

1. Pre-Sprint Planning

# Sprint Planning Checklist

## 1-2 Days Before Planning Meeting
- [ ] Groom product backlog (ensure top items are detailed)
- [ ] Update user story acceptance criteria
- [ ] Identify dependencies and blockers
- [ ] Prepare estimates from previous sprints
- [ ] Review team velocity (average story points per sprint)
- [ ] Identify team availability/absences
- [ ] Prepare sprint goals draft

## Information to Gather
- Product Owner priorities
- Team capacity (working hours available)
- Previous sprint metrics
- Upcoming holidays or interruptions
- Technical debt items to address

2. Sprint Planning Meeting Structure

// Example sprint planning agenda and execution

class SprintPlanner {
  constructor(team, sprintLength = 2) {
    this.team = team;
    this.sprintLength = sprintLength; // weeks
    this.userStories = [];
    this.sprintGoal = '';
    this.capacity = 0;
  }

  calculateTeamCapacity() {
    // Capacity = available hours - meetings - buffer
    const workHours = 40; // per person per week
    const meetingHours = 5; // estimated standups, retros, etc.
    const bufferPercent = 0.2; // 20% buffer for interruptions

    const capacityPerPerson = (workHours - meetingHours) * (1 - bufferPercent);
    this.capacity = capacityPerPerson * this.team.length * this.sprintLength;

    return this.capacity;
  }

  conductPlanningMeeting() {
    return {
      part1: {
        duration: '15 minutes',
        activity: 'Product Owner presents sprint goal',
        deliverable: 'Team understands business objective'
      },
      part2: {
        duration: '45-60 minutes',
        activity: 'Team discusses and estimates user stories',
        deliverable: 'Prioritized sprint backlog with story points'
      },
      part3: {
        duration: '15 minutes',
        activity: 'Team commits to sprint goal',
        deliverable: 'Formal sprint backlog committed'
      }
    };
  }

  createSprintBacklog(stories, capacity) {
    let currentCapacity = capacity;
    const sprintBacklog = [];

    for (let story of stories) {
      if (currentCapacity >= story.points) {
        sprintBacklog.push({
          ...story,
          status: 'planned',
          sprint: this.currentSprint
        });
        currentCapacity -= story.points;
      }
    }

    return {
      goal: this.sprintGoal,
      backlog: sprintBacklog,
      remainingCapacity: currentCapacity,
      utilization: ((capacity - currentCapacity) / capacity * 100).toFixed(1) + '%'
    };
  }
}

3. Story Point Estimation

# Story point estimation using Planning Poker approach

class StoryEstimation:
    # Fibonacci sequence for estimation
    ESTIMATE_OPTIONS = [1, 2, 3, 5, 8, 13, 21, 34]

    @staticmethod
    def calculate_story_points(complexity, effort, risk):
        """
        Estimate story points based on multiple factors
        Factors should be rated 1-5
        """
        base_points = (complexity * effort) / 5
        risk_multiplier = 1 + (risk * 0.1)
        estimated_points = base_points * risk_multiplier

        # Round to nearest Fibonacci number
        for estimate in StoryEstimation.ESTIMATE_OPTIONS:
            if estimated_points <= estimate:
                return estimate

        return StoryEstimation.ESTIMATE_OPTIONS[-1]

    @staticmethod
    def conduct_planning_poker(team_estimates):
        """
        Handle Planning Poker consensus process
        """
        estimates = sorted(team_estimates)
        median = estimates[len(estimates) // 2]

        # If significant disagreement, discuss and re-estimate
        if estimates[-1] - estimates[0] > 5:
            return {
                'consensus': False,
                'median': median,
                'low': estimates[0],
                'high': estimates[-1],
                'action': 'Discuss and re-estimate'
            }

        return {'consensus': True, 'estimate': median}

# Example usage
print(StoryEstimation.calculate_story_points(
    complexity=3,  # Medium complexity
    effort=2,      # Low effort
    risk=1         # Low risk
))  # Output: 3 points

4. Sprint Goal Definition

Sprint Goal Template:

Sprint: Sprint 23 (Nov 7 - Nov 20)

Goal Statement: |
  Enable users to manage multiple payment methods with a secure,
  intuitive interface that reduces checkout time by 40%

Success Criteria:
  - Payment method management UI implemented
  - 95% test coverage on payment logic
  - Performance: <200ms payment processing
  - Zero critical security issues
  - Feature released to 20% of users for A/B testing

Team Commitment: 89 story points
Expected Velocity: 85-95 points

Key Risks:
  - Payment gateway API changes
  - Security compliance requirements
  - Integration complexity

Acceptance:
  - All criteria met
  - Feature deployed to staging
  - Security review approved
  - Team and PO sign-off

5. Daily Standup Management

// Daily standup structure and tracking

class DailyStandup {
  constructor(team) {
    this.team = team;
    this.standups = [];
  }

  conductStandup(date) {
    const standup = {
      date,
      startTime: new Date(),
      participants: [],
      timeboxed: true,
      durationMinutes: 15
    };

    for (let member of this.team) {
      standup.participants.push({
        name: member.name,
        yesterday: member.getYesterdayWork(),
        today: member.getPlanForToday(),
        blockers: member.getBlockers(),
        helpNeeded: member.getHelpNeeded()
      });
    }

    return {
      standup,
      followUpActions: this.identifyFollowUps(standup),
      blockerResolutionOwners: this.assignBlockerOwners(standup)
    };
  }

  identifyFollowUps(standup) {
    return standup.participants
      .filter(p => p.blockers.length > 0)
      .map(p => ({
        owner: p.name,
        blockers: p.blockers,
        deadline: new Date(Date.now() + 24 * 60 * 60 * 1000)
      }));
  }
}

Best Practices

✅ DO

  • Base capacity on actual team velocity from past sprints
  • Include buffer time for interruptions and support work
  • Focus sprint goal on business value, not technical tasks
  • Timeboxe planning meeting (2 hours max for 2-week sprint)
  • Include entire team in planning discussion
  • Break down large stories into smaller, manageable pieces
  • Track story points for velocity trending
  • Review and adjust estimates based on actual completion
  • Maintain consistent sprint length
  • Include retrospective improvements in planning

❌ DON'T

  • Plan for 100% capacity utilization
  • Skip story grooming before planning meeting
  • Add stories after sprint starts (unless emergency)
  • Let one person estimate for entire team
  • Use story points as employee performance metrics
  • Ignore team velocity trends
  • Plan without clear sprint goal
  • Force stories into sprints to match capacity numbers
  • Skip sprint planning to save time
  • Use planning poker results as final estimate without discussion

Sprint Planning Tips

  • Keep stories to 5-13 points (break larger stories down)
  • Maintain 85-90% capacity utilization (leave buffer for interruptions)
  • Document sprint goal visibly (physical board or Jira)
  • Track velocity over 5-10 sprints to identify trends
  • Use historical data for better estimates
  • Celebrate sprint successes in reviews
  • Identify and address estimation biases
  • Adjust processes based on retrospective feedback

レビュー

まだレビューはありません。使ってみた感想をお寄せください。

同じリポジトリのスキル

概要と使いどころ

Comprehensive China A-share stock analysis covering fundamental analysis, technical analysis, policy impact assessment, and market-specific features (T+1 trading, price limits, northbound capital flow). Use when user asks about A股分析, Chinese mainland stocks, Shanghai/Shenzhen listed stocks, or needs analysis considering China market characteristics.

日本語の概要は準備中です。原文の説明を表示しています。

nicepkg/ai-workflow2852026年1月20日 更新

Screen and filter A-share stocks based on fundamental metrics, technical indicators, capital flow, and custom criteria. Support multiple screening strategies including value investing, growth investing, momentum trading, and dividend hunting. Use when user wants to find stocks meeting specific criteria like "低PE高ROE股票", "北向资金加仓股", "突破年线的股票".

日本語の概要は準備中です。原文の説明を表示しています。

nicepkg/ai-workflow2852026年1月20日 更新

Generate high-converting ad copy for Google Ads, Meta (Facebook/Instagram), LinkedIn, and TikTok. Creates multiple variations with A/B testing in mind. Includes headlines, descriptions, CTAs, and platform-specific formats. Use when creating ads, generating ad variations, or A/B testing ad copy.

日本語の概要は準備中です。原文の説明を表示しています。

nicepkg/ai-workflow2852026年1月20日 更新

Add new skills to an existing workflow and update all related documentation. Use when user wants to add skills from GitHub URLs to a workflow (e.g., "add this skill to the workflow", "为工作流添加技能"). Triggers on adding skills to workflows, updating workflow documentation after skill additions.

日本語の概要は準備中です。原文の説明を表示しています。

nicepkg/ai-workflow2852026年1月20日 更新

agile

無料

Agile product management, Scrum practices, and team collaboration for iterative product development.

日本語の概要は準備中です。原文の説明を表示しています。

nicepkg/ai-workflow2852026年1月20日 更新

Detects and removes AI writing patterns from text. Use when polishing drafts, cleaning AI-generated content, or ensuring writing sounds authentically human. Invoke with "run through slop detector", "clean this up", or "remove AI fingerprints".

日本語の概要は準備中です。原文の説明を表示しています。

nicepkg/ai-workflow2852026年1月20日 更新

nicepkg のスキルをすべて見る

このスキルの問題を報告する