cloud-agent analyze

The analyze command scans your project to detect the technology stack and recommend cloud services. By default it runs locally without any API key. With the --ai flag, it uses an AI agent for deeper analysis.

Usage

cloud-agent analyze [options]

Options

Flag	Description
--local	Force local-only analysis (no AI, no API key needed). This is the default behavior.
--ai	Enable AI-powered deep analysis using the analyzer agent. Requires an API key.
-p, --path <directory>	Path to the project directory to analyze. Defaults to the current working directory.

Local Analysis (Default)

When run without --ai, the analyze command performs a fast, offline scan of your project directory. It reads configuration files directly without calling any AI model.

What It Detects

Runtime detection by checking for:

• package.json — Node.js
• requirements.txt — Python
• go.mod — Go
• index.html (no other runtime) — Static HTML

Framework detection from dependencies:

• Node.js: Next.js, Express, NestJS, Fastify, React, Vue, Svelte, Astro, Hono
• Python: FastAPI, Flask, Django

Package manager by checking for lock files:

• pnpm-lock.yaml — pnpm
• yarn.lock — yarn
• bun.lockb — bun
• Default — npm (for Node.js) or pip (for Python)

Database detection from dependencies:

• PostgreSQL (pg, postgres, @prisma/client, psycopg2)
• MySQL (mysql, mysql2)
• MongoDB (mongodb, mongoose, pymongo)
• Redis (redis, ioredis)
• SQLite (sqlite3, better-sqlite3)

Docker detection:

• Dockerfile presence
• docker-compose.yml / docker-compose.yaml presence

Build and start commands from package.json scripts.

Port detection from script values (looks for PORT= patterns, defaults to 3000).

Quick Recommendations

Based on the detected project type, it suggests services:

Type	AWS	GCP	Azure
Static	S3 + CloudFront	Firebase Hosting	Static Web Apps
API	ECS Fargate or Lambda	Cloud Run	Container Apps or App Service
Web	Amplify or ECS Fargate	Cloud Run or App Engine	App Service or Static Web Apps
Container	ECS Fargate	Cloud Run	Container Apps

AI Analysis (--ai)

When run with --ai, the command invokes the analyzer agent — an AI agent powered by Gemini 2.0 Flash or GPT-4o Mini. The agent autonomously uses tools to:

1. Scan the project directory structure
2. Read configuration files
3. Analyze dependencies
4. Determine deployment strategy

The AI analysis provides:

• Detailed service recommendations for all three clouds
• Estimated monthly costs per provider
• Confidence score
• Environment variable requirements
• Deployment-specific recommendations

Requirements

• An AI API key must be configured (GOOGLE_GENERATIVE_AI_API_KEY or OPENAI_API_KEY)
• Run cloud-agent init if you need to set up a key

Example Output

  ── Project Analysis ──────────────────────

  ℹ Project: /home/user/my-express-app
  ℹ Mode: Local

  ℹ Project Details:

    Runtime:         Node.js
    Framework:       Express.js
    Project Type:    api
    Package Manager: npm
    Build Command:   tsc
    Start Command:   node dist/index.js
    Port:            3000
    Docker:          Yes
    Databases:       PostgreSQL

  Dependencies (8):
    - express
    - pg
    - dotenv
    - cors
    - helmet
    - zod
    - typescript
    - @types/express
    ... and 2 more

  ℹ Quick Recommendations:

    AWS:   ECS Fargate or Lambda
    GCP:   Cloud Run
    Azure: Container Apps or App Service

  ✓ Local analysis complete

  ℹ Tip: Run cloud-agent analyze --ai for AI-powered deep analysis

Exit Codes

Code	Meaning
0	Analysis completed successfully
1	Analysis failed (file read error, invalid project, etc.)