AI tools decision matrix

This reference helps you choose the right AI tool for a given task. Rather than defaulting to whichever tool you used last, use the matrix below to match your task to the tool best suited for it. The goal is efficiency: use lightweight tools for simple tasks, structured workflows for complex ones, and specialized plugins for domain-specific automation.

---

Decision matrix

The following table maps common platform team tasks to the recommended AI tool and explains why that tool is the best fit.

---

Decision flowchart

Use this flowchart when you are unsure which tool to reach for.

Start here: What are you trying to do?
|
+-- Is this a quick fix or small change?
|   |
|   +-- YES --> Claude Code CLI (no lifecycle overhead needed)
|   |
|   +-- NO --> Continue below
|
+-- Is this a new feature requiring design and requirements?
|   |
|   +-- YES --> AI-DLC + Claude Code CLI
|   |           (inception phase gathers requirements first)
|   |
|   +-- NO --> Continue below
|
+-- Do you need inline code suggestions while typing in VS Code?
|   |
|   +-- YES --> GitHub Copilot
|   |
|   +-- NO --> Continue below
|
+-- Do you need to work across multiple repositories simultaneously?
|   |
|   +-- YES --> GitHub Copilot Agent Mode
|   |           (see Multi-Repo Copilot Operations guide)
|   |
|   +-- NO --> Continue below
|
+-- Do you need to automate Teams or Outlook?
|   |
|   +-- YES --> Teams Automation or Outlook Automation plugin
|   |
|   +-- NO --> Continue below
|
+-- Are you looking for existing AI assets or plugins?
|   |
|   +-- YES --> Professor Hudak plugin (/ph-discover)
|   |
|   +-- NO --> Start with Claude Code CLI
|              (it handles most general tasks well)

---

Tool capabilities comparison

This matrix shows what each tool can and cannot do, so you can understand the tradeoffs.

Reading the comparison

• Claude Code CLI is the general-purpose workhorse. It handles most tasks well and is the default choice when no specialized tool is better suited.
• AI-DLC adds structure on top of Claude Code. Use it when the task is complex enough that skipping requirements or design would create risk.
• GitHub Copilot excels at real-time editing assistance and multi-repo awareness. It is complementary to Claude Code, not a replacement.
• Specialized plugins handle specific domains (communication, asset discovery, documentation) that the general tools do not address.

---

When NOT to use AI tools

Not every task benefits from AI assistance. Using AI tools for the following tasks adds overhead without proportional value:

The overhead threshold

A useful rule of thumb: if describing the task to the AI tool takes longer than doing the task yourself, skip the AI tool. AI assistance pays off when the task involves exploration, boilerplate generation, or working in unfamiliar territory.

---

Task-specific guidance

Ansible development

Terraform development

Documentation tasks

---

Related resources

• Claude Code Setup -- Installation and initial configuration
• AI-DLC Methodology -- Understanding the structured development lifecycle
• Plugin Catalog -- Complete inventory of available plugins with installation details
• AI Governance and Security -- Data sensitivity and compliance guidelines
• Multi-Repo Copilot Operations -- Using Copilot across the ohemr ecosystem
• Prompt Engineering for Infrastructure -- Crafting effective prompts for Ansible, Terraform, and Azure tasks