# Init

> Scan a project and generate cloudstack.json configuration. Use when setting up cloudstack-ai-plugins for the first time, onboarding a new project, or when the user says "init", "setup", "configure", "get started", or "onboard".

- **Type:** Skill
- **Install:** `agentstack add skill-makigjuro-cloudstack-ai-plugins-init`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [makigjuro](https://agentstack.voostack.com/s/makigjuro)
- **Installs:** 0
- **Category:** [Cloud & Infrastructure](https://agentstack.voostack.com/c/cloud-infrastructure)
- **Latest version:** 0.1.0
- **License:** MIT
- **Upstream author:** [makigjuro](https://github.com/makigjuro)
- **Source:** https://github.com/makigjuro/cloudstack-ai-plugins/tree/main/plugins/dev-workflow/skills/init
- **Website:** https://github.com/makigjuro/cloudstack-ai-plugins#quick-start

## Install

```sh
agentstack add skill-makigjuro-cloudstack-ai-plugins-init
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

# Initialize Project

Scan the project structure, detect the tech stack, and generate a `cloudstack.json` configuration file. This is the first thing users should run after installing cloudstack-ai-plugins.

## Arguments

- `--minimal` -- Generate only the essential fields, skip optional sections

## Process

### Step 1: Check for existing config

```bash
test -f cloudstack.json && echo "EXISTS" || echo "NONE"
```

If `cloudstack.json` already exists, ask the user whether to overwrite or update it.

### Step 2: Detect project structure

Run these scans in parallel:

**Backend detection:**
```bash
# .NET solution
find . -name "*.sln" -maxdepth 3 -not -path "./.git/*" 2>/dev/null

# Shared domain project
find . -path "*/Domain/Entities" -type d -not -path "*/bin/*" -not -path "*/obj/*" 2>/dev/null

# Shared infrastructure project
find . -path "*Shared*Infrastructure*" -name "*.csproj" -not -path "*/bin/*" 2>/dev/null

# Microservices (directories containing .Application subfolder)
find . -name "*.Application" -type d -not -path "*/bin/*" -not -path "*/obj/*" 2>/dev/null

# Messaging framework
grep -rl "WolverineFx\|MediatR\|MassTransit" --include="*.csproj" -l 2>/dev/null | head -1

# ORM detection
grep -rl "Microsoft.EntityFrameworkCore" --include="*.csproj" -l 2>/dev/null | head -1
```

**Frontend detection:**
```bash
# Package.json location
find . -name "package.json" -maxdepth 3 -not -path "*/node_modules/*" -not -path "*/tools/*" 2>/dev/null

# If found, detect framework and libraries from dependencies
cat /package.json | grep -E "react|vue|angular|@tanstack|zustand|redux|shadcn|@mui|vite|webpack|next"
```

**Infrastructure detection:**
```bash
# Terraform
find . -name "*.tf" -maxdepth 4 -not -path "./.git/*" 2>/dev/null | head -1

# Terragrunt
find . -name "terragrunt.hcl" -maxdepth 4 2>/dev/null | head -1

# Helm charts
find . -name "Chart.yaml" -maxdepth 4 2>/dev/null

# Docker Compose
find . -name "docker-compose*.yml" -o -name "docker-compose*.yaml" -maxdepth 3 2>/dev/null

# .NET Aspire
find . -name "*AppHost*" -type d -maxdepth 4 2>/dev/null

# CI/CD
ls .github/workflows/*.yml 2>/dev/null || ls .gitlab-ci.yml 2>/dev/null || ls azure-pipelines.yml 2>/dev/null
```

**Repository detection:**
```bash
git remote get-url origin 2>/dev/null
```

### Step 3: Resolve ambiguities

Use AskUserQuestion for anything that can't be auto-detected:

- **Multiple solutions found** -- Ask which is the primary one
- **No namespace detectable** -- Ask for the root C# namespace
- **Multiple frontend directories** -- Ask which is the main one
- **Cloud provider unclear** -- Ask Azure vs AWS vs GCP

Keep questions to a minimum. If a value is clearly detectable, don't ask.

### Step 4: Generate cloudstack.json

Build the configuration object from detected values. Include:

- `$schema` reference for IDE autocompletion
- `project` -- name (from solution name or directory), namespace (from .csproj)
- `repository` -- owner/name (from git remote)
- `backend` -- solutionPath, sharedDomainPath, sharedInfraPath, services[], orm, messaging
- `frontend` -- path, framework, stateManagement, uiLibrary, bundler, devPort
- `infrastructure` -- cloud, iac, iacWrapper, terraformPath, chartsPath, namespace
- `localDev` -- orchestrator, appHostPath or composePath
- `conventions` -- branchFormat (detect from existing branches), commitStyle

If `--minimal` was passed, only include `project`, `backend.solutionPath`, and `backend.services[]`.

Write the file:
```bash
# Write cloudstack.json to project root
```

### Step 5: Recommend plugins

Based on what was detected, recommend which plugins to install:

```markdown
## Project Initialized

Created `cloudstack.json` with:
- {N} microservices detected
- {frontend framework} frontend at `{path}/`
- {iac tool} infrastructure at `{path}/`

### Recommended plugins:

| Plugin | Why | Install |
|--------|-----|---------|
| dotnet-architect | {N} .NET services detected | `/plugin install dotnet-architect@cloudstack-ai-plugins` |
| react-developer | React frontend at `{path}/` | `/plugin install react-developer@cloudstack-ai-plugins` |
| cloud-infra | Terraform + Helm detected | `/plugin install cloud-infra@cloudstack-ai-plugins` |
| dev-workflow | Git workflow + CI/CD | `/plugin install dev-workflow@cloudstack-ai-plugins` |

### Next steps:
1. Review `cloudstack.json` and adjust any values
2. Install recommended plugins above
3. Try `/dotnet-architect:check-architecture` to validate your project
```

Only recommend plugins relevant to what was detected (e.g., skip react-developer if no frontend found).

## Error Handling

- **Not a git repo:** Warn but continue -- skip repository section
- **No .NET solution found:** Skip backend section, note it
- **No frontend found:** Skip frontend section
- **Permission denied on scan:** Note the directory and skip it

## Related Skills

- `/dotnet-architect:check-architecture` -- Run after init to validate architecture
- `/dev-workflow:start-work` -- Begin implementing with proper branch workflow

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [makigjuro](https://github.com/makigjuro)
- **Source:** [makigjuro/cloudstack-ai-plugins](https://github.com/makigjuro/cloudstack-ai-plugins)
- **License:** MIT
- **Homepage:** https://github.com/makigjuro/cloudstack-ai-plugins#quick-start

Install and usage instructions live in the source repository linked above.

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** no
- **Filesystem access:** no
- **Shell / process execution:** no
- **Environment & secrets:** no
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/skill-makigjuro-cloudstack-ai-plugins-init
- Seller: https://agentstack.voostack.com/s/makigjuro
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.