From project-orchestrator
Generate CLAUDE.md, settings.json, hooks, subagents, commands, and MCP config. Auto-detects tech stack from project files. Use when setting up or managing the development workspace and tooling.
How this skill is triggered — by the user, by Claude, or both
Slash command
/project-orchestrator:agent-workspace-setupThis skill is limited to the following tools:
The summary Claude sees in its skill listing — used to decide when to auto-load this skill
Set up project workspaces for agent-orchestrated microservices development.
Set up project workspaces for agent-orchestrated microservices development.
mkdir -p services/{api-gateway/src/{common,config,modules},core-service/src/{modules,common,database},core-service/prisma,ai-service/app/{api,services,models,tasks,integrations},ai-service/tests,shared/{proto,schemas}}
mkdir -p apps/{web/src/{app,components,lib,hooks,types},mobile-flutter/lib/{features,core,app},mobile-flutter/test,mobile-kmp/{shared,androidApp,iosApp}}
mkdir -p infrastructure/{docker,terraform,k8s,scripts} docs .github/workflows
Detect the stack by reading manifest files in each service directory:
[ -f services/core-service/package.json ] && grep -q "@nestjs/core" services/core-service/package.json && echo "NestJS"
[ -f services/ai-service/requirements.txt ] && grep -q "django" services/ai-service/requirements.txt && echo "Django"
[ -f apps/web/package.json ] && grep -q "next" apps/web/package.json && echo "Next.js"
[ -f apps/mobile-flutter/pubspec.yaml ] && echo "Flutter"
[ -f apps/mobile-kmp/shared/build.gradle.kts ] && echo "KMP"
Generate from detected stack. Must include: Project Overview, Tech Stack, Build & Run commands per service, Test Commands, Lint & Format commands, and Rules (violations are bugs). See the root CLAUDE.md for the canonical format.
{
"permissions": {
"allow": [
"Bash(npm run *)", "Bash(npx *)", "Bash(pytest *)",
"Bash(docker-compose *)", "Bash(flutter *)",
"Bash(git status)", "Bash(git diff *)", "Bash(git log *)",
"Read", "Write", "Edit", "Grep", "Glob"
],
"deny": ["Bash(rm -rf /*)"]
}
}
{
"hooks": [
{ "event": "PreCommit", "command": "npm run lint && npm run format:check", "workingDirectory": "services/core-service" },
{ "event": "PreCommit", "command": "ruff check . && ruff format --check .", "workingDirectory": "services/ai-service" }
]
}
# .env.example — include all service configs with safe placeholders
API_GATEWAY_PORT=3000
CORE_SERVICE_PORT=3001
DATABASE_URL=postgresql://postgres:postgres@localhost:5432/core_db
REDIS_URL=redis://localhost:6379
AI_SERVICE_PORT=8000
DJANGO_SECRET_KEY=change-me
ANTHROPIC_API_KEY=
node_modules/
dist/
.env
.env.local
*.pyc
__pycache__/
.venv/
.dart_tool/
build/
.next/
coverage/
.idea/
git init
git add -A
git commit -m "chore: initial project scaffold
- Monorepo structure with services and apps
- CLAUDE.md with project conventions
- Docker Compose for local development
- Environment configuration templates"
npx claudepluginhub vivekmano27/agent-orchestrator --plugin project-orchestratorGuides project setup through 6 phases: detects tech stack from package.json/requirements.txt/etc., creates CLAUDE.md, configures MCP memory and auto-loop hooks. For new/existing Claude Code projects.
Bootstraps a repository with claude-leverage stack conventions: AGENTS.md, CLAUDE.md, .gitignore patterns, and optional language-specific logging templates. Interactive and idempotent.
Bootstraps .claude/ dotclaude config from template if missing, then customizes all files to match project's tech stack, conventions, and patterns. Interactive via user confirmations.