Claude Code Plugins

Community-maintained marketplace

Feedback

|

Install Skill

1Download skill
2Enable skills in Claude

Open claude.ai/settings/capabilities and find the "Skills" section

3Upload to Claude

Click "Upload skill" and select the downloaded ZIP file

Note: Please verify skill by going through its instructions before using it.

SKILL.md

name monorepo-team-lead
description Principal-level team leadership and management skill for monorepos. This skill should be used when designing ownership structures, managing teams, routing tasks between humans and AI, and handling cross-team coordination. Use when: setting up CODEOWNERS, onboarding team members, making task routing decisions, writing RFCs, or coordinating cross-domain work. Essential for AI Manager capabilities.

Monorepo Team Lead

Overview

This skill provides expert-level (L4) capabilities for team management, code ownership, human-AI collaboration, and cross-team coordination in monorepo environments.

Skill Level: Principal (E5/L4) Matrix Coverage: T.* (Team), M.* (Manager Skills)

Code Ownership

CODEOWNERS Template

# .github/CODEOWNERS

# Domain ownership
/domains/auth/           @team-auth
/domains/content/        @team-content
/domains/ai/             @team-ai

# Platform ownership
/platform/               @team-platform
/infrastructure/         @team-sre

# Critical paths (require senior)
*.lock                   @team-platform
/platform/auth/          @team-auth @security-leads

# AI layer
/.claude/                @team-ai @team-devex

Ownership Principles

1. Every file has exactly ONE owner
2. Owners MUST approve PRs to their areas
3. Cross-cutting changes require ALL owners
4. No orphan code (everything is owned)

Human-AI Task Routing

Decision Matrix

HUMAN ONLY:
- Architecture decisions
- Breaking change approval
- Security-critical changes
- Cross-team coordination
- Production deployments

AI-ASSISTED (Human reviews):
- Feature implementation
- Test writing
- Documentation
- Refactoring
- Dependency updates

AI AUTONOMOUS:
- Lint fixes
- Formatting
- Type error fixes
- Simple bug fixes

Routing Decision Tree

Task arrives:
├── Strategic? → HUMAN MANAGER
├── Needs judgment? → HUMAN ENGINEER
├── Well-defined & repeatable? → AI AGENT
├── Cross-domain? → HUMAN COORDINATES
└── Default:
    ├── Simple → AI (with review)
    ├── Complex → HUMAN (AI assists)
    └── Production → HUMAN required

Onboarding

Week 1 Checklist

## Day 1-2: Setup
- [ ] Clone and install: pnpm install
- [ ] Verify build: pnpm build
- [ ] Understand structure: ls -la

## Day 3-4: First PR
- [ ] Find "good first issue"
- [ ] Create branch, make change
- [ ] Submit PR, respond to review

## Day 5: Learn System
- [ ] Read CODEOWNERS
- [ ] Understand CI pipeline
- [ ] Know who to ask

Knowledge Transfer

KT CHECKLIST:
[ ] Domain architecture explained
[ ] Key files/modules identified
[ ] Common tasks demonstrated
[ ] Gotchas/pitfalls documented
[ ] Contacts for help listed

RFC Process

RFC Template

# RFC: [Title]

## Summary
One paragraph explaining the proposal.

## Motivation
Why? What problem does it solve?

## Design
Technical details.

## Affected Packages
- @myorg/pkg-a: [changes]
- @myorg/pkg-b: [changes]

## Migration
How will existing code migrate?

## Timeline
- Week 1: Implementation
- Week 2: Migration

## Approvals
- [ ] @team-owner
- [ ] @team-affected

Cross-Team Coordination

Protocol

1. CREATE RFC
   - Describe change
   - List affected domains
   - Propose migration

2. GET ALIGNMENT
   - Tag all owners
   - Get written approval

3. IMPLEMENT IN PHASES
   - Add new (no breaking)
   - Migrate consumers
   - Remove old

4. COMMUNICATE
   - Announce in channels
   - Update documentation

Manager Competencies

AI Work Evaluation

EVALUATE AI OUTPUT:
- Does it match requirements?
- Is code subtly wrong?
- Are edge cases handled?
- Is it over-engineered?
- Did it respect boundaries?

Quality Gates

Gate AI Human Manager
Tests pass Must Must Review
Breaking change Flag Review Approve
Cross-domain Flag Coordinate Approve
Security Pass Verify Approve

Quick Reference

# Ownership
cat .github/CODEOWNERS | grep <path>

# RFC
gh issue create --template=rfc.md

# Onboarding
cat ONBOARDING.md

# Task routing
# Strategic → Human Manager
# Judgment → Human Engineer
# Repeatable → AI Agent