⚡ FrootAI CLI

Scaffold projects, search knowledge, estimate costs, and validate configs — all from the terminal.
Ships with the MCP Server package. No extra install needed.

No install needed — runs via npx

npx frootai <command>

Or if you've installed frootai-mcp globally: just frootai <command>

📋 Commands

🚀 `frootai init`

Interactive project scaffolding. 3 questions → complete FrootAI project.

$ npx frootai init

  🌳 FrootAI™ CLI v3.1.2
  From the Roots to the Fruits

  Let's set up your FrootAI-powered project!

  What are you building?
  1) Enterprise RAG (document Q&A, knowledge base)
  2) AI Agent (multi-step reasoning, tool calling)
  3) AI Gateway (API management, cost control)
  4) Content Moderation (safety, compliance)
  5) Multi-modal (vision, document intelligence)
  6) Custom (pick from 20 solution plays)

  Choose [1-6]: 1

  Target scale?
  1) dev   — Local development, minimal cost
  2) prod  — Production, high availability

  Choose [1-2]: 1

  Project name [my-ai-project]: my-rag-app

What gets created

my-rag-app/
├── .vscode/mcp.json          ← MCP server auto-connects in VS Code
├── .github/
│   ├── agents/               ← Builder, Reviewer, Tuner agent definitions
│   ├── copilot-instructions.md
│   └── instructions/
├── config/
│   ├── openai.json           ← Model, temperature, max_tokens
│   ├── search.json           ← Azure AI Search settings
│   └── guardrails.json       ← Safety + PII + grounding rules
├── evaluation/
│   └── eval-config.json      ← Groundedness, relevance thresholds
├── infra/                    ← Infrastructure templates
└── README.md

Then: cd my-rag-app && code . — FrootAI MCP auto-connects, agents are ready.

📐 `frootai scaffold <play>`

One-command play scaffold. Creates all 5 FROOT kits + froot.json manifest.

$ npx frootai scaffold 01-enterprise-rag
# or shorthand:
$ npx frootai scaffold play-01

  🌳 FrootAI™ CLI v3.1.2

  Scaffolding play: 01-enterprise-rag

  ✅ Play scaffolded: 01-enterprise-rag

  Files created:
  .vscode/mcp.json             ← MCP auto-connect
  .github/agents/              ← Builder + Reviewer + Tuner
  .github/instructions/        ← 5 WAF instruction files
  config/                      ← OpenAI + Search + Guardrails
  evaluation/                  ← Eval config + thresholds
  spec/play-spec.json          ← SpecKit (WAF alignment)
  froot.json                   ← Play manifest (5 kits)

Auto-detects existing projects and merges files. Includes froot.json manifest linking all 5 kits.

🔍 `frootai search <query>`

Search the FrootAI knowledge base (16 FROOT modules) from the terminal.

$ npx frootai search "RAG architecture"

  Searching: "RAG architecture"

  Found 3 result(s):

  📖 R2 — RAG Architecture & Retrieval
     Layer: Reasoning
  📖 O4 — Azure AI Platform & Landing Zones
     Layer: Orchestration
  📖 T3 — Production Architecture Patterns
     Layer: Transformation

💰 `frootai cost [play] --scale <dev|prod>`

Estimate Azure infrastructure costs for any solution play at dev or prod scale.

$ npx frootai cost enterprise-rag --scale prod

  Cost Estimate: enterprise-rag (prod)

  Service                        | Cost
  ───────────────────────────────┼─────────────────────────
  AI Search                      | $250/mo (S1)
  OpenAI (GPT-4o)                | ~$150/mo (moderate)
  App Service                    | $55/mo (B1)
  Storage                        | $0.02/GB/mo
  Cosmos DB                      | $25/mo (400 RU/s)
  Total estimate                 | ~$500-800/mo

✅ `frootai validate`

Check that your project has the right structure, configs, and MCP setup.

$ npx frootai validate

  ✅ MCP config — .vscode/mcp.json
  ✅ Copilot instructions — .github/copilot-instructions.md
  ✅ Builder agent — .github/agents/builder.agent.md
  ✅ Reviewer agent — .github/agents/reviewer.agent.md
  ✅ Tuner agent — .github/agents/tuner.agent.md
  ✅ OpenAI config — config/openai.json
  ✅ Guardrails config — config/guardrails.json
  ✅ MCP server configured: frootai

  8/8 checks passed

🩺 `frootai doctor`

Health check for your development environment — verifies Node.js, npm, git, VS Code, and MCP connectivity.

$ npx frootai doctor

  ✅ Node.js v22.0.0 (>= 18 required)
  ✅ npm 10.8.0
  ✅ git version 2.45.0
  ✅ VS Code 1.96.0
  ✅ .vscode/mcp.json found
  ✅ .github/agents/ found
  ✅ frootai-mcp available on npm

📌 `frootai version`

$ npx frootai version
frootai-mcp v3.1.2

🔗 How It Connects

┌──────────────────────────────────────────────┐
│  npx frootai init                            │
│  ├── Scaffolds project structure             │
│  ├── Creates .vscode/mcp.json               │
│  ├── Creates .github/agents/ (3 agents)      │
│  └── Creates config/ (3 config files)        │
└──────────────┬───────────────────────────────┘
               │
               ▼
┌──────────────────────────────────────────────┐
│  code .                                      │
│  └── VS Code reads .vscode/mcp.json         │
│      └── Auto-starts frootai-mcp server      │
│          └── 22 tools available to Copilot   │
└──────────────┬───────────────────────────────┘
               │
               ▼
┌──────────────────────────────────────────────┐
│  @builder / @reviewer / @tuner               │
│  └── Copilot agents use MCP tools            │
│      └── Architecture patterns, cost est,    │
│          config validation, play search      │
└──────────────────────────────────────────────┘

⚡ FrootAI CLI

📋 Commands

🚀 frootai init

📐 frootai scaffold <play>

🔍 frootai search <query>

💰 frootai cost [play] --scale <dev|prod>

✅ frootai validate

🩺 frootai doctor

📌 frootai version