Scrum Conductor
Overview
Facilitates AI-enhanced Scrum orchestration with automated ticket management and high-velocity sprint coordination. Synthesizes daily updates from git activity, detects blockers proactively, and maintains backlog integrity across issue trackers.
When to use: Sprint planning, daily standups, backlog grooming, ticket creation, velocity forecasting, sprint retrospectives, estimation, risk management, release planning, cross-tracker synchronization.
When NOT to use: Technical implementation tasks (use specialized coding skills), architecture design, security auditing.
Quick Reference
| Pattern | Approach | Key Points |
|---|---|---|
| Fact-first standups | Auto-generate from git logs and PRs | Never ask humans for data in commit history |
| Ticket engineering | Machine-readable DoD with acceptance criteria | Binary true/false criteria, implementation pointers |
| Sprint planning | Capacity calculation with focus factor and velocity | Budget = avg velocity (last 3 sprints) adjusted for absences |
| Estimation | Planning Poker with Fibonacci scale, T-shirt sizing for roadmap | Reference stories anchor the scale, re-calibrate quarterly |
| Capacity forecasting | Historical cycle time and lead time | Factor holidays, context debt, and bottlenecks |
| Backlog grooming | AI clustering and deduplication | Flag tickets older than 2 sprints for archive |
| Backlog refinement | 1-2 sessions per sprint, Definition of Ready checklist | Stories must meet DoR before entering sprint planning |
| Blocker detection | Scan PRs, assignments, and dependencies | Flag stale PRs, OOO assignees, breaking deps |
| Parking lot | Move deep-dives out of standups | Standups focus on status and blockers only |
| Sprint retrospective | Start/Stop/Continue with action item tracking | Track retro action completion rate across sprints |
| Risk management | Dependency mapping, risk register (probability x impact) | Address risk scores 6+ immediately, monitor 3-5 |
| Release planning | Multi-sprint roadmap with confidence levels | Feature flags decouple deployment from release |
| Escalation tiers | Tier 1 autonomous, Tier 2 clarification, Tier 3 pairing | Match response to complexity |
| Priority frameworks | MoSCoW, value vs effort matrix, WSJF | WSJF favors small high-value items |
Conductor Protocol
- Ceremony Initialization: Identify the current sprint phase (Planning, Daily, Review, Retro)
- Telemetry Sync: Pull recent activity from git commits, PRs, and communication channels
- Fact Synthesis: Generate factual summaries before ceremonies begin
- Verification: Confirm all action items are converted into tracked tickets with clear owners and DoD
Common Mistakes
| Mistake | Correct Pattern |
|---|---|
| Asking developers for status updates available in commit logs | Auto-generate fact summaries from PRs and merges before standups |
| Creating tickets without a machine-readable Definition of Done | Every ticket needs explicit acceptance criteria and technical pointers |
| Guessing sprint velocity without historical data | Use cycle time, holidays, and context debt to forecast capacity |
| Letting standups run beyond 15 minutes with deep-dive discussions | Move deep-dives to a parking lot session; standups focus on blockers |
| Allowing the backlog to grow to 200+ items without pruning | Auto-flag tickets older than 2 sprints for archive or refactor |
| Using AI to replace human conversation | Use AI to prepare for the conversation, not substitute it |
| Ignoring team sentiment and morale signals | High velocity with low morale is a leading indicator of burnout |
| No Definition of Done agreed by the team | Establish a team-wide DoD checklist applied to every story |
| Skipping retrospectives when the sprint "went fine" | Every sprint has improvement opportunities; consistency builds the habit |
| Sprint planning without capacity calculation | Calculate capacity: members x available days x focus factor |
| Stories entering sprint without acceptance criteria | Enforce Definition of Ready before stories enter a sprint |
| Estimating stories individually instead of as a team | Use Planning Poker so the full team contributes perspective |
| Slicing stories horizontally by layer | Slice vertically through all layers so each ticket delivers working functionality |
| No dependency mapping between sprint stories | Map dependencies explicitly and identify the critical path |
| Retro action items with no owner or due date | Every action item needs an owner, a due date, and follow-up tracking |
Delegation
- Synthesize daily standup summaries from git activity and PRs: Use
Taskagent to pull commit logs and generate fact-based updates - Cluster and deduplicate backlog tickets across issue trackers: Use
Exploreagent to scan GitHub Issues, Jira, and Linear for similar or conflicting items - Plan sprint capacity and risk assessment using historical velocity: Use
Planagent to model delivery probability and identify at-risk items
References
- Automated daily rituals, fact-checking workflows, blocker detection, sprint planning, retrospectives, and refinement
- Ticket engineering standards, user story templates, estimation techniques, and priority frameworks
- Predictive velocity, sprint metrics, risk management, distributed teams, and release planning
- Multi-agent task handoffs, escalation tiers, AI retrospectives, blocker detection, and human escalation guardrails