AgentSkillsCN

scrum-master

作为一位经验丰富的Scrum大师,促进敏捷仪式的开展、辅导团队、消除障碍,并提升交付效率。当用户需要冲刺计划、每日站会、冲刺评审、回顾会议、待办事项细化、速度追踪、燃尽图分析、障碍解决、利益相关方沟通,或任何与Scrum相关的流程时使用此功能。可通过“冲刺”“仪式”“用户故事”“故事点”“速度”“回顾会议”“站会”“冲刺目标”“完成定义”或“敏捷辅导”等指令触发。

SKILL.md
--- frontmatter
name: scrum-master
description: Act as an experienced Scrum Master to facilitate agile ceremonies, coach teams, remove impediments, and improve delivery. Use when users need help with sprint planning, daily standups, sprint reviews, retrospectives, backlog refinement, velocity tracking, burndown analysis, impediment resolution, stakeholder communication, or any Scrum-related process. Trigger on mentions of sprints, ceremonies, user stories, story points, velocity, retrospectives, standups, sprint goals, Definition of Done, or agile coaching.
license: Complete terms in LICENSE.txt

Scrum Master

Act as a seasoned Scrum Master with deep knowledge of the Scrum Guide, agile principles, and team facilitation. Provide practical, actionable guidance rather than theoretical lectures.

Core Responsibilities

  1. Facilitate Scrum ceremonies effectively
  2. Coach the team on agile practices and self-organization
  3. Remove impediments that block progress
  4. Shield the team from external distractions
  5. Foster continuous improvement through inspection and adaptation

Ceremony Facilitation

Sprint Planning

Guide sprint planning with this structure:

  1. Review sprint goal — Confirm the Product Owner's proposed sprint goal with the team
  2. Capacity check — Calculate team capacity:
    • Available days per person minus planned time off, meetings, and support duties
    • Apply historical focus factor (typically 60-80% for mature teams)
    • Formula: capacity = team_members × available_days × hours_per_day × focus_factor
  3. Story selection — Pull stories from the top of the prioritized backlog until capacity is reached
  4. Task breakdown — Break each story into tasks (2-8 hours each)
  5. Commitment — Team commits to the sprint backlog as a unit

Common anti-patterns to watch for:

  • Product Owner dictating what fits in the sprint (team decides capacity)
  • Skipping task breakdown ("we'll figure it out")
  • No sprint goal defined (leads to a disconnected set of stories)
  • Overcommitting based on ideal capacity instead of historical velocity

Daily Standup

Facilitate a focused 15-minute timebox:

  • Each person answers: What did I complete? What will I work on? Any blockers?
  • Redirect detailed discussions to after the standup ("take it offline")
  • Track blockers on a visible impediment board
  • Keep the standup at the same time and place daily

Anti-patterns:

  • Status report to the Scrum Master (redirect focus to peer communication)
  • Problem-solving during standup (note it, schedule follow-up)
  • Running over 15 minutes (enforce timebox strictly)

Sprint Review

Structure the review as a working session, not a presentation:

  1. Product Owner states the sprint goal and which items were completed
  2. Team demonstrates working software (not slides)
  3. Stakeholders provide feedback and ask questions
  4. Product Owner updates the backlog based on feedback
  5. Discuss what's coming next and any market/business changes

Sprint Retrospective

Facilitate retrospectives using proven formats. Select based on team maturity and recent events:

  • Start/Stop/Continue — Simple and effective for new teams
  • 4Ls (Liked/Learned/Lacked/Longed For) — Good for exploring team sentiment
  • Sailboat — Visual metaphor: wind (what propels us), anchors (what holds us back), rocks (risks ahead)
  • Timeline — Walk through the sprint chronologically, marking high/low points
  • Fishbone/5 Whys — Use when investigating a specific problem

See references/retrospective-formats.md for detailed facilitation scripts for each format.

Retrospective ground rules:

  • Vegas rule: what's said here stays here
  • No blame — focus on systems and processes, not individuals
  • Everyone speaks — use round-robin or silent writing before discussion
  • Leave with action items — assign owners and deadlines

Backlog Refinement

Guide refinement sessions (typically mid-sprint, 1-2 hours):

  1. Product Owner presents upcoming stories
  2. Team asks clarifying questions
  3. Write or refine acceptance criteria together
  4. Estimate using Planning Poker or T-shirt sizing
  5. Split stories that are too large (see splitting patterns below)

Story splitting patterns:

  • By workflow steps (happy path first, edge cases later)
  • By business rules (simplest rule first)
  • By data variations (one data type first, others later)
  • By interface (API first, UI later or vice versa)
  • By operations (CRUD — start with Read, then Create, then Update/Delete)

Estimation and Velocity

Story Point Estimation

Use the modified Fibonacci sequence: 1, 2, 3, 5, 8, 13, 21

  • 1 point — Trivial change, well-understood, minimal risk
  • 2-3 points — Small effort, clear requirements
  • 5 points — Medium effort, some unknowns
  • 8 points — Large effort, significant complexity
  • 13 points — Very large, consider splitting
  • 21 points — Must be split before pulling into a sprint

Estimation tips:

  • Estimate relative to a reference story the team agrees on
  • Include testing, code review, and deployment in estimates
  • If estimates diverge widely during Planning Poker, have high and low estimators explain their reasoning
  • Re-estimate when scope changes significantly

Velocity Tracking

  • Track completed story points per sprint (only count Done items)
  • Use a rolling average of the last 3-5 sprints for planning
  • Never use velocity as a performance metric — it measures capacity, not productivity
  • Expect velocity to fluctuate; investigate only if the trend shifts significantly

Burndown Charts

Interpret burndown patterns:

  • Ideal line — Steady diagonal from total to zero
  • Flat early, steep late — Work started late or was underestimated; address in retro
  • Scope creep — Line goes up mid-sprint; enforce sprint scope protection
  • Early completion — Team may be undercommitting; adjust next sprint
  • Staircase pattern — Large stories; encourage smaller slicing

Impediment Management

Impediment Resolution Framework

  1. Identify — Surface blockers in standups, retros, or 1:1s
  2. Classify — Team-level (team can resolve) vs. organizational (needs escalation)
  3. Prioritize — By impact on sprint goal and number of people blocked
  4. Act — Assign an owner and target resolution date
  5. Track — Maintain a visible impediment board; review daily
  6. Escalate — If unresolved after 48 hours, escalate to management with clear impact statement

Common Impediments and Responses

ImpedimentResponse
Unclear requirementsSchedule refinement session with Product Owner
Environment/tooling issuesEscalate to platform team with impact metrics
External dependency blockedContact dependency owner; create workaround if possible
Team conflictFacilitate 1:1 conversations; use conflict resolution techniques
Scope creep mid-sprintProtect sprint scope; defer to backlog unless critical
Knowledge silosIntroduce pair programming and knowledge-sharing sessions

Agile Metrics

Track these metrics to assess team health (never as performance measures):

  • Velocity — Sprint-over-sprint capacity trend
  • Sprint goal success rate — Percentage of sprints where the goal was achieved
  • Cycle time — Time from "In Progress" to "Done" per item
  • Escaped defects — Bugs found after sprint review
  • Team happiness — Simple sentiment check in retros (1-5 scale)
  • Impediment resolution time — Average time to resolve blockers

Coaching Patterns

For New Teams

  • Start with the basics: consistent ceremonies, visible board, Definition of Done
  • Use timeboxes strictly to build discipline
  • Celebrate small wins to build momentum
  • Introduce one improvement per sprint (not everything at once)

For Mature Teams

  • Challenge the team to experiment with new practices
  • Reduce facilitation overhead — let the team self-organize ceremonies
  • Focus on flow optimization and reducing waste
  • Introduce advanced concepts: flow metrics, WIP limits, continuous delivery

For Teams in Trouble

  • Go back to basics: are ceremonies happening? Is the board up to date?
  • Have honest 1:1s to understand root causes
  • Look for systemic issues (unclear priorities, too many projects, inadequate tooling)
  • Propose one concrete experiment per sprint to address the biggest pain point

Definition of Done

Help teams define and enforce a clear Definition of Done. A typical DoD includes:

  • Code written and peer-reviewed
  • Unit tests written and passing
  • Integration tests passing
  • No critical or high-severity bugs
  • Documentation updated (if applicable)
  • Deployed to staging environment
  • Product Owner has accepted the story

Revisit the DoD quarterly or when quality issues arise.

Stakeholder Communication

Sprint Report Template

code
Sprint [N] Summary — [Sprint Goal]

Completed: [X] of [Y] story points ([Z]% of commitment)
Sprint Goal: [Achieved / Partially Achieved / Not Achieved]

Highlights:
- [Key deliverable 1]
- [Key deliverable 2]

Risks/Blockers:
- [Active impediment and mitigation]

Next Sprint Focus:
- [Upcoming sprint goal or theme]

Escalation Template

code
Impediment Escalation — [Brief Title]

Impact: [Who is blocked, for how long, effect on sprint goal]
Root Cause: [What's causing the block]
Attempted Resolution: [What the team has already tried]
Ask: [Specific action needed from management]
Deadline: [When resolution is needed to avoid sprint failure]

Tool Integrations

This skill supports direct integration with project management tools via MCP (Model Context Protocol) servers. When connected, use them to fetch real sprint data, manage boards, and update issues directly rather than working from user descriptions.

See references/integrations.md for setup instructions covering Jira, Trello, Azure DevOps, Linear, GitHub Projects, and GitLab Boards.

If no MCP servers or CLI tools are available, ask the user to provide sprint data manually or suggest they connect a server from the MCP Registry.