Sprint Planner

Sprint Planner Skill

Purpose

Plan and initialize a new sprint by selecting approved specifications from the project board, creating a milestone, and generating atomic GitHub issues for each task. This skill bridges the gap between approved specs and executable work items, breaking down high-level features into trackable, atomic tasks for the development team.

When to Use

Use this skill in the following situations:

• Starting a new sprint or development cycle
• Converting approved specs into actionable GitHub issues
• Creating a milestone and task breakdown for a feature
• Initializing work items from the project board backlog
• Breaking down an epic into atomic implementation tasks

Prerequisites

• Project board configured with "Approved Backlog" status
• Approved spec files exist in docs/specs/ directory
• Specs have task lists with - [ ] checkbox format
• GitHub repository set up with project board
• gh CLI tool installed and authenticated
• jq tool installed for JSON parsing
• Epic issues linked to spec files in issue body

Workflow

Step 1: Review Project Board

Check the project board for approved specs ready to be planned:

gh project item-list PROJECT_NUMBER --owner @me --format json

Identify epics in the "Approved Backlog" status. These represent specs that have been reviewed, approved, and are ready for implementation.

Step 2: Discuss Sprint Scope with User

Engage the user to determine:

• Which epic(s) to include in the sprint
• Sprint timeline and duration
• Sprint goals and priorities
• Resource availability

Review the available epics together and select which work to prioritize for this sprint.

Step 3: Define Sprint Metadata

Work with the user to establish:

• Sprint name: e.g., "Sprint 4", "Q1 2025 Sprint 2"
• Sprint goals: High-level objectives for this development cycle
• Timeline: Start and end dates (if applicable)
• Success criteria: How to measure sprint completion

Step 4: Run the Helper Script

Execute the sprint planning script to automate GitHub issue creation:

bash scripts/create-sprint-issues.sh

Step 5: Understand What the Script Does

The helper script automates these steps:

1. Queries project board:

   • Fetches all items in "Approved Backlog" status
   • Displays available epics with their issue numbers

2. Prompts for epic selection:

   • User enters the epic issue number to plan
   • Script retrieves epic details and body

3. Extracts spec file:

   • Parses epic body to find associated spec file path
   • Validates the spec file exists

4. Prompts for milestone name:

   • User enters sprint milestone name
   • Script creates the GitHub milestone

5. Parses tasks from spec:

   • Reads spec file for task list items (- [ ] **Title**: Description)
   • Creates a GitHub issue for each task

6. Creates GitHub issues:

   • Each task becomes an issue with "TASK: " prefix
   • Issues reference the spec file and parent epic
   • Issues are assigned to the sprint milestone

Step 6: Verify Issue Creation

After the script completes:

gh issue list --milestone "Sprint 4"

Review that:

• All expected tasks are present
• Issue titles are clear and actionable
• Issues reference correct spec and epic
• Milestone assignment is correct

Step 7: Review Created Issues with User

Walk through the created issues together:

• Confirm task breakdown is complete
• Verify priorities are set appropriately
• Discuss any missing tasks or adjustments needed
• Assign issues if team members are identified

Step 8: Update Project Board (Optional)

If using project board automation:

• Move epic from "Approved Backlog" to "In Sprint"
• Verify task issues appear on project board
• Set priorities or labels as needed

Step 9: Document Sprint Plan (Optional)

Consider creating a sprint planning document:

# Sprint 4 Plan

**Timeline**: Nov 1 - Nov 15, 2025
**Epic**: #44 - Claude Code Skill Compliance
**Milestone**: Sprint 4

## Goals
- Restructure all 7 SynthesisFlow skills
- Improve Claude Code compliance
- Validate with skill validation script

## Tasks
- #45 - Restructure doc-indexer skill
- #46 - Restructure project-init skill
...

Error Handling

jq Not Installed

Symptom: Script reports jq command not found

Solution:

• Install jq: sudo apt install jq (Linux)
• Or: brew install jq (Mac)
• Verify: jq --version

No Approved Epics Found

Symptom: Script reports no epics in approved backlog

Solution:

• Verify project board has "Approved Backlog" status
• Check if epics are in correct status column
• Confirm specs have been approved via spec PR
• Run spec-authoring skill to create and approve specs first

Epic Body Missing Spec Reference

Symptom: Script cannot find spec file in epic body

Solution:

• Update epic issue body to include spec file path
• Format: docs/specs/feature-name/spec.md
• Ensure spec file actually exists at that path
• Follow spec-authoring workflow to create spec if needed

Milestone Already Exists

Symptom: GitHub API returns error that milestone exists

Solution:

• Use a different milestone name
• Or manually delete existing milestone if unused
• Or add tasks to existing milestone via gh issue create

Spec File Missing Task List

Symptom: Script creates no issues or very few issues

Solution:

• Verify spec file has task list in correct format: - [ ] **Title**: Description
• Ensure tasks are in a clear section (e.g., ## Tasks or ## Implementation)
• Review spec-delta.md template for proper task formatting
• Update spec file to include detailed task breakdown

Permission Denied Creating Milestone

Symptom: GitHub API returns 403 or permission error

Solution:

• Verify gh CLI is authenticated: gh auth status
• Ensure you have repository write access
• Re-authenticate if needed: gh auth login
• Verify repository name is correct in script REPO variable

Configuration Notes

The script uses these configuration variables (lines 6-11):

PROJECT_NUMBER="1"
APPROVED_BACKLOG_ID="888736cd"
OWNER="@me"
REPO="bodangren/git-workflow"

To adapt for your project:

1. Update PROJECT_NUMBER to your GitHub project number
2. Find APPROVED_BACKLOG_ID by querying project field values
3. Update REPO to your repository path (owner/repo)
4. Keep OWNER="@me" to use current authenticated user

Finding your Approved Backlog ID:

gh project field-list PROJECT_NUMBER --owner @me

Notes

• Script automates repetitive work: Creating dozens of issues manually is tedious and error-prone
• LLM guides the strategic decisions: Which specs to plan, sprint scope, timeline
• Task format matters: Spec files must use - [ ] **Title**: Description format for parsing
• One epic per sprint run: Run the script multiple times if planning multiple epics
• Milestone groups tasks: All tasks from same epic share a milestone for easy filtering
• Epic remains parent: Each task issue references parent epic for traceability
• Project board integration: Issues can be automatically added to project board if configured
• Idempotent considerations: Running script twice creates duplicate issues - be careful
• Manual refinement expected: Review and adjust created issues as needed after script runs
• Sprint planning is collaborative: Engage user throughout the process for priorities and scope