when-managing-github-projects-use-github-project-management

GitHub Project Management Skill

Overview

Automate and orchestrate GitHub project management workflows using intelligent agent coordination. This skill provides comprehensive project management capabilities including automated issue triage, sprint planning, milestone tracking, project board synchronization, and team coordination through GitHub Projects v2 integration.

When to Use This Skill

Activate this skill when planning and executing development sprints, managing issue backlogs and triage queues, coordinating work across distributed teams, automating project board updates and status tracking, tracking milestones and release schedules, or establishing project management workflows for new teams.

Use for both small team projects (2-5 developers) and large-scale coordination (10+ team members), agile/scrum sprint management, kanban-style continuous delivery, or hybrid project management approaches.

Agent Coordination Architecture

Swarm Topology

Initialize a star topology with a central coordinator agent managing specialized agents for issue tracking, planning, and board synchronization. Star topology enables centralized control with efficient communication to specialized agents.

# Initialize star swarm for project management
npx claude-flow@alpha swarm init --topology star --max-agents 6 --strategy balanced

Specialized Agent Roles

Central Coordinator (coordinator): Hub agent that orchestrates all project management activities. Makes prioritization decisions, coordinates between teams, and ensures consistency across project artifacts. Acts as project manager.

Planner (planner): Handles sprint planning, capacity estimation, and resource allocation. Creates sprint goals, breaks down epics into stories, estimates effort, and balances workload across team members.

Issue Tracker (issue-tracker): Automates issue triage, labeling, assignment, and lifecycle management. Monitors new issues, applies initial classification, routes to appropriate team members, and tracks resolution progress.

Project Board Sync (project-board-sync): Maintains GitHub Projects v2 boards, updates issue status, manages custom fields, and enforces workflow automation. Ensures boards accurately reflect current project state.

Project Management Workflows (SOP)

Workflow 1: Automated Issue Triage

Process incoming issues with intelligent classification and routing.

Phase 1: Issue Ingestion

Step 1.1: Initialize Star Topology

# Set up star swarm with coordinator hub
mcp__claude-flow__swarm_init topology=star maxAgents=6 strategy=balanced

# Spawn coordinator and specialists
mcp__claude-flow__agent_spawn type=coordinator name=coordinator
mcp__claude-flow__agent_spawn type=researcher name=issue-tracker
mcp__claude-flow__agent_spawn type=researcher name=planner
mcp__claude-flow__agent_spawn type=coordinator name=project-board-sync

Step 1.2: Monitor New Issues

Set up real-time monitoring for new issues:

# Subscribe to new issues
bash scripts/github-webhook.sh subscribe \
  --repo <owner/repo> \
  --events "issues" \
  --callback "scripts/process-new-issue.sh"

Or use MCP real-time subscription if Flow-Nexus available:

mcp__flow-nexus__realtime_subscribe table=issues event=INSERT

Step 1.3: Fetch Issue Details

When new issue created, fetch comprehensive context:

# Get issue details with comments and reactions
bash scripts/github-api.sh fetch-issue \
  --repo <owner/repo> \
  --issue <number> \
  --include-comments true

Phase 2: Intelligent Classification

Step 2.1: Analyze Issue Content

Task("Issue Tracker", "
  Analyze and classify new issue #<NUMBER>:

  1. Read issue body and comments from memory: github/issues/<NUMBER>
  2. Classify issue type: bug|feature|enhancement|documentation|question
  3. Determine severity: critical|high|medium|low
  4. Identify affected components using references/component-map.md
  5. Estimate complexity: simple|moderate|complex
  6. Check for duplicates using similarity search
  7. Extract key entities: affected files, error messages, user impact

  Use scripts/issue-classifier.sh for ML-based classification
  Store classification in memory: github/triage/<NUMBER>
  Run hooks: npx claude-flow@alpha hooks pre-task --description 'issue triage'
", "issue-tracker")

Step 2.2: Apply Labels and Metadata

Based on classification, apply appropriate labels:

# Apply automated labels
bash scripts/github-api.sh add-labels \
  --repo <owner/repo> \
  --issue <number> \
  --labels "<type>,<severity>,<component>"

# Set custom fields in project board
bash scripts/project-api.sh set-field \
  --project <project-id> \
  --issue <number> \
  --field "Complexity" \
  --value "<complexity>"

Step 2.3: Assign Team Member

Route to appropriate team member based on expertise:

Task("Coordinator", "
  Assign issue #<NUMBER> to appropriate team member:

  1. Review team expertise map in references/team-skills.md
  2. Check current workload using scripts/team-capacity.sh
  3. Match issue requirements with team member skills
  4. Consider time zone and availability
  5. Assign issue via GitHub API

  Store assignment decision in memory: github/assignments/<NUMBER>
", "coordinator")

# Assign issue
bash scripts/github-api.sh assign-issue \
  --repo <owner/repo> \
  --issue <number> \
  --assignee <username>

Phase 3: Priority and Milestone Assignment

Step 3.1: Determine Priority

Task("Planner", "
  Prioritize issue #<NUMBER> within backlog:

  1. Assess business impact (user-facing, revenue, compliance)
  2. Evaluate technical risk (security, stability, data loss)
  3. Consider dependencies (blocking other work?)
  4. Review stakeholder input from issue comments
  5. Calculate priority score using references/priority-matrix.md

  Store priority in memory: github/priorities/<NUMBER>
", "planner")

Step 3.2: Assign to Milestone

Based on priority and current sprint capacity:

# Add to appropriate milestone
bash scripts/github-api.sh set-milestone \
  --repo <owner/repo> \
  --issue <number> \
  --milestone "<sprint-name>"

Step 3.3: Update Project Board

Task("Project Board Sync", "
  Add issue #<NUMBER> to project board:

  1. Determine appropriate project board (team, initiative, sprint)
  2. Add issue to board in 'Triage' column
  3. Set custom fields: Priority, Complexity, Sprint, Team
  4. Link related issues and PRs
  5. Add to sprint view if assigned to active sprint

  Use scripts/project-api.sh for Projects v2 API
  Store board update in memory: github/boards/<NUMBER>
", "project-board-sync")

Workflow 2: Sprint Planning Automation

Plan and execute development sprints with agent coordination.

Phase 1: Sprint Preparation

Step 1.1: Analyze Team Capacity

# Calculate available capacity for next sprint
bash scripts/team-capacity.sh calculate \
  --team <team-name> \
  --start-date <YYYY-MM-DD> \
  --duration-days 14 \
  --include-pto true \
  --output references/sprint-capacity.json

Step 1.2: Review Backlog

Task("Planner", "
  Prepare sprint planning meeting:

  1. Fetch all backlog items using scripts/github-api.sh list-issues
  2. Filter by priority (high and critical first)
  3. Identify dependencies between issues
  4. Group related issues into themes
  5. Create preliminary sprint proposal based on capacity

  Reference references/sprint-capacity.json for team availability
  Store sprint proposal in memory: github/sprint/proposal
", "planner")

Phase 2: Sprint Composition

Step 2.1: Create Sprint Milestone

# Create new sprint milestone
bash scripts/github-api.sh create-milestone \
  --repo <owner/repo> \
  --title "Sprint <number>: <theme>" \
  --due-date <YYYY-MM-DD> \
  --description "Sprint goals: ..."

**Step 2.2: Assig