devops-plan

Gather DevOps context for infrastructure planning. Detects IaC tools, providers, and recommends skills. Use when /majestic:plan detects infrastructure work.

⭐13stars🔀0forks📁View on GitHub🕐Updated Jan 3, 2026⬇️2downloads

When & Why to Use This Skill

The devops-plan skill is a specialized Claude agent tool designed to automate the discovery and analysis of infrastructure context. It scans project environments to identify Infrastructure as Code (IaC) tools such as Terraform, OpenTofu, and Ansible, detects specific cloud providers, and performs preliminary security audits. By providing structured findings and recommending specialized sub-skills, it significantly accelerates the planning phase of DevOps workflows and ensures architectural consistency.

Use Cases

  • Automated Infrastructure Discovery: Quickly scan a repository to identify existing Terraform, Ansible, or Cloud-init configurations without manual inspection.
  • Cloud Provider Identification: Automatically detect which cloud platforms (AWS, GCP, Azure, Hetzner, etc.) are being used in a project to ensure the correct deployment tools are invoked.
  • Pre-Implementation Security Auditing: Run automated security reviews on detected IaC files to identify potential vulnerabilities before infrastructure changes are executed.
  • Agentic Workflow Orchestration: Enable AI agents to intelligently select the most appropriate specialized coding skills based on the detected technology stack and providers.
namedevops-plan
descriptionGather DevOps context for infrastructure planning. Detects IaC tools, providers, and recommends skills. Use when /majestic:plan detects infrastructure work.
allowed-toolsRead, Grep, Glob, Bash, Task

DevOps Plan Context

Gather infrastructure context for planning. Returns structured findings for the parent plan command.

When Called

This skill is invoked by /majestic:plan when DevOps keywords are detected in the feature description.

Workflow

1. Detect IaC Files

Check for existing infrastructure code:

# Terraform/OpenTofu
ls *.tf infra/*.tf terraform/*.tf infrastructure/*.tf 2>/dev/null

# Ansible
ls ansible.cfg playbook.yml ansible/*.yml playbooks/*.yml 2>/dev/null
ls -d roles/ ansible/roles/ 2>/dev/null

# Cloud-init
ls cloud-init*.yml user-data*.yml cloud-config*.yml 2>/dev/null

Set flags:

  • HAS_TERRAFORM: true/false
  • HAS_ANSIBLE: true/false
  • HAS_CLOUD_INIT: true/false

2. Detect Providers

If Terraform files exist, extract providers:

grep -h "provider\s*\"" *.tf infra/*.tf 2>/dev/null | grep -oE '"[a-z]+"' | tr -d '"' | sort -u

Common providers to detect:

  • aws, google, azurerm - Major clouds
  • hcloud - Hetzner
  • digitalocean - DigitalOcean
  • cloudflare - Cloudflare
  • backblaze - Backblaze B2

3. Map Skills

Based on detection, recommend skills:

Detection Skill Purpose
Any .tf files majestic-devops:opentofu-coder HCL patterns, state management
Provider: hcloud majestic-devops:hetzner-coder Hetzner-specific resources
Provider: cloudflare majestic-devops:cloudflare-coder Cloudflare zones, DNS, Workers
Provider: backblaze majestic-devops:backblaze-coder B2 storage patterns
Ansible files majestic-devops:ansible-coder Playbook patterns
Cloud-init files majestic-devops:cloud-init-coder VM provisioning

4. Security Review

If existing IaC files found, run security audit:

Task(subagent_type: "majestic-devops:infra-security-review", prompt: "Audit existing infrastructure code for security issues")

Capture findings for the plan.

5. Return Context

Return structured context to the calling plan command:

## Infrastructure Context

### Detected Tools
- **Terraform/OpenTofu:** [yes/no] - [version if found]
- **Ansible:** [yes/no]
- **Cloud-init:** [yes/no]

### Providers
[List detected providers]

### Skills to Apply
| Skill | Purpose |
|-------|---------|
| `majestic-devops:opentofu-coder` | [reason] |
| ... | ... |

### Security Findings
[Summary from infra-security-review or "No existing IaC to audit"]

### Recommendations
- [Any specific recommendations based on detection]

Notes

  • This skill gathers context only - no implementation
  • Skills are recommendations, not requirements
  • Security review only runs if IaC files exist