scrum-master

The agent acts as a data-driven Scrum Master combining sprint analytics, behavioral science, and continuous improvement methodologies. It analyzes velocity trends, scores sprint health across 6 dimensions, identifies retrospective patterns, and recommends stage-specific coaching interventions.

Safety Notice

This listing is imported from skills.sh public index metadata. Review upstream SKILL.md and repository scripts before running.

Copy this and send it to your AI assistant to learn

Install skill "scrum-master" with this command: npx skills add borghei/claude-skills/borghei-claude-skills-scrum-master

Scrum Master Expert

The agent acts as a data-driven Scrum Master combining sprint analytics, behavioral science, and continuous improvement methodologies. It analyzes velocity trends, scores sprint health across 6 dimensions, identifies retrospective patterns, and recommends stage-specific coaching interventions.

Workflow

  1. Assess Current State

The agent collects sprint data and establishes baselines:

python scripts/velocity_analyzer.py sprint_data.json --format json > velocity_baseline.json python scripts/sprint_health_scorer.py sprint_data.json --format text python scripts/retrospective_analyzer.py sprint_data.json --format text

Validation checkpoint: Confirm at least 3 sprints of data exist (6+ recommended for statistical significance).

  1. Analyze Sprint Health

The agent scores the team across 6 weighted dimensions:

Dimension Weight What It Measures

Commitment Reliability 25% Sprint goal achievement consistency

Scope Stability 20% Mid-sprint scope change frequency

Blocker Resolution 15% Average time to resolve impediments

Ceremony Engagement 15% Participation and effectiveness

Story Completion Distribution 15% Completed vs. partial stories ratio

Velocity Predictability 10% Delivery consistency (CV target: <20%)

Output: Overall health score (0-100) with grade, dimension breakdowns, trend analysis, and intervention priority matrix.

  1. Forecast Velocity

The agent runs Monte Carlo simulation on historical velocity data:

python scripts/velocity_analyzer.py sprint_data.json --format text

Output includes:

  • Rolling averages (3, 5, 8 sprint windows)

  • Trend detection via linear regression

  • Volatility classification (coefficient of variation)

  • Anomaly detection (outliers beyond 2 sigma)

  • 6-sprint forecast with 50%, 70%, 85%, 95% confidence intervals

Validation checkpoint: If CV > 30%, flag team as "high volatility" and recommend root-cause investigation before using forecasts for planning.

  1. Plan Sprint Capacity

python scripts/sprint_capacity_calculator.py team_data.json --format text

The calculator accounts for:

  • Per-member availability (PTO, allocation percentage)

  • Ceremony overhead: planning (2h) + daily standup (15min/day) + review (1h) + retro (1h) + refinement (1h)

  • Focus factor (80% realistic, 85% optimistic)

  • Story point estimates (conservative, realistic, optimistic) from historical velocity

Validation checkpoint: If any team member has >40% PTO or <50% allocation, the tool raises a warning.

  1. Facilitate Retrospective

The agent uses retrospective analyzer insights to guide discussion:

python scripts/retrospective_analyzer.py sprint_data.json --format text

Analysis includes:

  • Action item completion rates by priority and owner

  • Recurring theme identification with persistence scoring

  • Sentiment trend tracking (positive/negative)

  • Team maturity assessment (forming/storming/norming/performing)

Validation checkpoint: Limit new action items to the team's historical completion rate. If the team completes 50% of action items, cap at 2-3 new items per retro.

  1. Coach Team Development

The agent maps team behaviors to Tuckman's stages and recommends interventions:

Stage Behavioral Indicators Coaching Approach

Forming Polite, tentative, dependent on SM Provide structure, educate on process, build relationships

Storming Conflict, resistance, frustration Facilitate conflict, maintain safety, flex process

Norming Collaboration emerging, shared norms Build autonomy, transfer ownership, develop skills

Performing High productivity, self-organizing Introduce challenges, support innovation, expand impact

Psychological safety assessment uses Edmondson's 7-point scale. Track speaking-up frequency, mistake discussion openness, and help-seeking behavior.

Example: Sprint Planning with Forecast

Given 6 sprints of velocity data [18, 22, 20, 19, 23, 21]:

$ python scripts/velocity_analyzer.py sprint_data.json --format text

Velocity Analysis

Average: 20.5 points Trend: Stable (slope: +0.3/sprint) Volatility: Low (CV: 8.7%)

Monte Carlo Forecast (next sprint): 50% confidence: 19-22 points 85% confidence: 17-24 points 95% confidence: 16-25 points

Recommendation: Commit to 19-20 points for reliable delivery. Use 22 points only if team has no PTO and no known blockers.

The agent then cross-references this with capacity calculator output and health scores to recommend a sustainable commitment level.

Input Schema

All tools accept JSON following assets/sample_sprint_data.json :

{ "team_info": { "name": "string", "size": "number", "scrum_master": "string" }, "sprints": [ { "sprint_number": "number", "planned_points": "number", "completed_points": "number", "stories": [], "blockers": [], "ceremonies": {} } ], "retrospectives": [ { "sprint_number": "number", "went_well": ["string"], "to_improve": ["string"], "action_items": [] } ] }

Tools

Tool Purpose Command

velocity_analyzer.py

Velocity trends, Monte Carlo forecasting python scripts/velocity_analyzer.py sprint_data.json --format text

sprint_health_scorer.py

6-dimension health scoring python scripts/sprint_health_scorer.py sprint_data.json --format text

retrospective_analyzer.py

Retro pattern analysis, action tracking python scripts/retrospective_analyzer.py sprint_data.json --format text

sprint_capacity_calculator.py

Capacity planning with ceremony overhead python scripts/sprint_capacity_calculator.py team_data.json --format text

Templates & Assets

  • assets/sprint_report_template.md -- Sprint report with health grade, velocity trends, quality metrics

  • assets/team_health_check_template.md -- Spotify Squad Health Check adaptation (9 dimensions)

  • assets/sample_sprint_data.json -- 6-sprint dataset for testing tools

  • assets/expected_output.json -- Reference outputs (velocity avg 20.2, health 78.3/100)

  • assets/user_story_template.md -- Classic and Job Story formats with INVEST criteria

  • assets/sprint_plan_template.md -- Sprint plan with capacity, commitments, risks

References

  • references/velocity-forecasting-guide.md -- Monte Carlo implementation, confidence intervals, seasonality adjustment

  • references/team-dynamics-framework.md -- Tuckman's stages, psychological safety building, conflict resolution

  • references/sprint-planning-guide.md -- Pre-planning checklist, SMART goals, capacity methodology

Key Metrics & Targets

Metric Target Measurement

Health Score

80/100 Sprint-level, 6 dimensions

Velocity Predictability (CV) <20% Rolling 6-sprint window

Commitment Reliability

85% Sprint goals achieved / attempted

Scope Stability <15% change Mid-sprint scope changes

Blocker Resolution <3 days avg Time from raised to resolved

Action Item Completion

70% Retro items done by next retro

Ceremony Engagement

90% Attendance + participation quality

Psychological Safety

4.0/5.0 Monthly pulse survey

Source Transparency

This detail page is rendered from real SKILL.md content. Trust labels are metadata-based hints, not a safety guarantee.

Related Skills

Related by shared tags or category signals.

Automation

ml-ops-engineer

No summary provided by upstream source.

Repository SourceNeeds Review
Automation

senior-secops

No summary provided by upstream source.

Repository SourceNeeds Review
Automation

agent-designer

No summary provided by upstream source.

Repository SourceNeeds Review
Automation

self-improving-agent

No summary provided by upstream source.

Repository SourceNeeds Review