Skip to main content

In this article

Stage 5: Sprint Planning

Overview

Sprint Planning organizes decomposed work items into actionable sprints. This stage covers backlog triage, work discovery, priority assignment, and sprint scoping using tracker-agnostic backlog management tools.

When You Enter This Stage

You enter Sprint Planning after completing Stage 4: Decomposition with work items created and ready for prioritization.

NOTE

Prerequisites: Work items exist in GitHub Issues or ADO. Repository has labels and milestones configured for sprint tracking.

Available Tools

ToolTypeHow to InvokePurpose
backlog-managerAgentSelect backlog-manager agentManage the backlog end-to-end across ADO, GitHub, and Jira
backlog-plan discoverSkill/backlog-plan discoverFind candidate work items for sprint planning, read-only
backlog-plan triageSkill/backlog-plan triageClassify and label unprocessed items, read-only
backlog-plan sprintSkill/backlog-plan sprintCreate a sprint plan from backlog priorities, read-only
backlog-execute runSkill/backlog-execute runApply a reviewed handoff to the tracker
backlog-execute addSkill/backlog-execute addAdd a single new item to the backlog
backlog-managementSkillAuto-loaded by the Backlog ManagerSupplies backlog planning, triage, and story quality conventions

Role-Specific Guidance

TPMs lead Sprint Planning, balancing priorities across the backlog and coordinating with Tech Leads on technical sequencing. Tech Leads contribute capacity estimates and identify dependency chains.

Starter Prompts

Issue Discovery

Survey the backlog for candidate work. Discovery is read-only and writes an analysis file you review before continuing:

/backlog-plan discover

The workflow resolves your tracker, runs its preflight, and asks which discovery path fits. Supply search terms such as authentication and a target iteration such as v2.4.0 when prompted, or point it at a requirements document such as docs/architecture/prd-notifications.md.

Backlog Triage

Classify untriaged items with label suggestions, iteration assignment, and duplicate detection:

/backlog-plan triage

Triage is read-only. It writes a triage plan with its recommendations, and a separate execution pass applies them.

Sprint Planning

Build a prioritized sprint plan from a target iteration with capacity constraints and a sprint goal:

/backlog-plan sprint

Backlog Execution

Dry-run a reviewed handoff to preview item operations before committing changes:

/backlog-execute run --dry-run

When the preview matches your intent, apply it:

/backlog-execute run

Create a single new item conversationally, using repository templates where they exist:

/backlog-execute add

User Story Coaching

Select backlog-manager agent to create a new story from a rough idea. The agent applies the story quality guidance in the backlog-management skill's story-quality reference:

I need a story for adding webhook notifications when deployment status changes. The platform team needs real-time alerts in their monitoring dashboard.

Select backlog-manager agent to refine a vague existing story:

Help me refine this story: Title: Improve error handling, Description: Handle errors better, AC: Errors are handled

Full Backlog Orchestration

Select backlog-manager agent to coordinate triage and sprint planning end-to-end:

Prepare the v2.4.0 milestone for sprint planning. Triage any needs-triage issues first, then build a prioritized sprint plan with a 15-issue capacity.

Stage Outputs and Next Stage

Sprint Planning produces a scoped sprint with prioritized issues, assigned owners, and milestone targets. Transition to Stage 6: Implementation when the sprint is planned and work items are assigned.

🤖 Crafted with precision by ✨Copilot following brilliant human instruction, then carefully refined by our team of discerning human reviewers.