What is zMesh?

zMesh is an AI-first Backend as a Service that gives you a full backend — PostgreSQL database, serverless edge functions, file storage, AI gateway, authentication — without writing a single line of infrastructure code.

This CLI lets you manage everything from your terminal.

pip install zmesh-cli

Quick Start

# Authenticate
zmesh login

# Create a new project
zmesh init --name "My App"

# Create a table
zmesh database query "CREATE TABLE users (
  id SERIAL PRIMARY KEY,
  name TEXT NOT NULL,
  email TEXT UNIQUE
)"

# Deploy an edge function
zmesh functions deploy ./api.js

# Upload a file
zmesh storage upload ./logo.png --path assets/logo.png

That's it. Your backend is live.

Features

zmesh database tables
zmesh database query "SELECT * FROM users"
zmesh database migrate ./schema.sql
zmesh database seed ./data.json

zmesh functions deploy ./handler.js
zmesh functions invoke my-func --payload '{}'
zmesh functions logs my-func

zmesh storage upload ./image.png
zmesh storage list --prefix photos/
zmesh storage download photos/img.png

# Manage via Dashboard or API
# Chat, Embeddings, Vector Store (pgvector)
# Schema & RLS generation from prompts

Commands

Authentication & Projects

Database

Edge Functions

Deploy with options:

zmesh functions deploy ./send-email.js \
  --name send-email \
  --description "Transactional email sender" \
  --entry-point handler \
  --env SMTP_HOST=smtp.example.com \
  --timeout 10000

Storage

Environment Variables

Edge Functions

zMesh's edge functions run in a custom-built sandboxed runtime — no AWS Lambda, no third-party FaaS. Each function gets isolated subprocess execution with OS-level resource limits and a pre-injected zmesh SDK with direct database access.

JavaScript

// get-users.js
exports.handler = async (payload, zmesh) => {
  const users = await zmesh.db.query(
    'SELECT * FROM users WHERE active = $1', [true]
  );

  const newLog = await zmesh.db.insert('audit_log', {
    action: 'list_users',
    count: users.length
  });

  const secret = zmesh.env.API_SECRET;

  return { users, logged: true };
};

Python

# get-users.py
def handler(payload, zmesh):
    users = zmesh.db.query(
        "SELECT * FROM users WHERE active = %s", (True,)
    )

    zmesh.db.insert("audit_log", {
        "action": "list_users",
        "count": len(users)
    })

    return {"users": users}

SDK Reference

End-to-End Example

Build a complete blog API in under 2 minutes:

# 1. Create project
zmesh init --name "Blog API"

# 2. Create schema
zmesh database query "
CREATE TABLE posts (
  id SERIAL PRIMARY KEY,
  title TEXT NOT NULL,
  body TEXT,
  author TEXT,
  created_at TIMESTAMP DEFAULT NOW()
)"

# 3. Write a function
cat > get-posts.js << 'EOF'
exports.handler = async (payload, zmesh) => {
  const posts = await zmesh.db.query(
    'SELECT * FROM posts ORDER BY created_at DESC LIMIT $1',
    [payload.limit || 10]
  );
  return { posts };
};
EOF

# 4. Deploy
zmesh functions deploy ./get-posts.js

# 5. Test
zmesh functions invoke get-posts --payload '{"limit": 5}'

# 6. Call from your app
curl -X POST https://api.zmesh.in/api/functions/get-posts \
  -H "x-api-key: zb_your_key" \
  -H "Content-Type: application/json" \
  -d '{"limit": 5}'

Configuration

Self-Hosted

zmesh login --api-url https://api.yourdomain.com/v1

Requirements

• Python 3.11+
• A zMesh account

Links