Arodonata

A high-performance, async-first Python library for Check Point security management operations with intelligent PostgreSQL caching and automatic session handling.

Key Features • Installation • Quick Start • Architecture • Documentation

---

🚀 Overview

Arodonata — Async, Reliable Operations — is a production-grade Python library designed to simplify, accelerate, and scale automation against Check Point Security Management Servers.

Standard Check Point API client scripts often struggle with connection overhead, rate limits, session expiration, and slow response times when dealing with tens of thousands of security objects. Arodonata solves these bottlenecks by wrapping operations in an async-first engine coupled with an intelligent database-backed cache layer.

Why Arodonata?

• ⚡ High-Concurrency Execution: Fully asynchronous execution using aiohttp and asyncio, letting you execute thousands of queries simultaneously.
• 🔐 Zero-Config Session Handling: Transparent authentication, session pooling, and auto-recovery/re-login if a session expires.
• 💾 Intelligent DB Caching: Blazing-fast local caching powered by PostgreSQL (with JSONB support) and SQLAlchemy.
• 🔄 Smart Refresh (show-changes): Avoid full cache rebuilds. The cache tracks and pulls only incremental modifications from Check Point management logs.
• 🛡️ Strict Type-Safety: 100% type hints validated with Pydantic v2 models, catching structural errors before runtime.

---

✨ Key Features

Feature	Description	Benefit
Async Core	Fully non-blocking network calls	Scales to enterprise deployments with zero thread overhead
Session Cache	Automatic database-backed session token persistence	Bypasses repetitive logins, preserving firewall resources
Multi-Domain (MDM)	Native domain resolution and context propagation	Operates across complex tenant environments seamlessly
Rate Limiting	Active per-server request gating and database locks	Prevents overloading firewalls during large-scale tasks
SSE Event Streams	Live Server-Sent Events for background refresh tasks	Real-time monitoring of sync status and progress bars
MCP Server	Streamable-HTTP MCP endpoint with the Check Point show_* tool surface	Lets Claude and other MCP clients query policy through the cache

---

📦 Installation

Arodonata requires Python 3.13+ and a PostgreSQL 12+ database for caching.

Choose the package extra that fits your project:

• arodonata: Core library for Python automation, scripts, and applications (high-performance async client, session pooling, PostgreSQL caching, and declarative CPCRUD engine).
• arodonata[mcp]: Core library + Model Context Protocol (MCP) server support, including the arodonata-mcp CLI executable and arodonata.mcp ASGI integration for FastAPI/Starlette.

Adding to Your Project

With uv (Recommended)

# Core library
uv add arodonata

# OR with MCP server support
uv add "arodonata[mcp]"

With Standard pip

# Core library
pip install arodonata

# OR with MCP server support
pip install "arodonata[mcp]"

In pyproject.toml

[project]
dependencies = [
    # Core library for Check Point automation & caching:
    "arodonata>=1.11.0",

    # OR if you need the MCP server / embedded ASGI tools:
    # "arodonata[mcp]>=1.11.0",
]

In requirements.txt

arodonata>=1.11.0
# or
arodonata[mcp]>=1.11.0

Local Development / Contributing

git clone https://github.com/chkp-antonr/arodonata.git
cd arodonata
uv sync --dev

---

⚡ Quick Start

1. Set Up Your Environment

Create a .env file containing your database cache connection details and Check Point server coordinates:

# Database Configuration (PostgreSQL with JSONB recommended)
DATABASE_URL=postgresql+asyncpg://cp_user:cp_password@localhost:5432/arodonata_cache

# Management Server Details (supports multiple servers)
MGMT_NAMES=primary-mgmt,backup-mgmt
MGMT_SERVERS=192.168.10.10,192.168.10.11
API_KEY_VARS=PRIMARY_MGMT_KEY,BACKUP_MGMT_KEY

# Secrets (resolved at runtime from environment / secrets manager)
PRIMARY_MGMT_KEY=your-primary-api-key-here
BACKUP_MGMT_KEY=your-backup-api-key-here

2. Using arodonata in Python Projects

Use ArodonataClient directly in your application. The calling application manages the database engine lifecycle and passes the engine into the client:

import asyncio
import os
from sqlalchemy.ext.asyncio import create_async_engine
from arodonata import ArodonataClient, ArodonataSettings

async def main():
    # 1. Resolve database URL & configuration settings
    database_url = os.getenv("DATABASE_URL", "sqlite+aiosqlite:///:memory:")
    engine = create_async_engine(database_url)
    
    settings = ArodonataSettings(
        mgmt_names=os.getenv("MGMT_NAMES", "mgmt_srv"),
        mgmt_servers=os.getenv("MGMT_SERVERS", "10.0.0.1"),
        api_keys=os.getenv("PRIMARY_MGMT_KEY", "mock-key"),
    )

    # 2. Initialize the client
    client = ArodonataClient(engine=engine, settings=settings)

    try:
        # 3. Enter async context manager for automatic session management
        async with client:
            servers = client.get_mgmt_names()
            print(f"Connected to management nodes: {servers}")

            # Query hosts from the primary server (read from cache or live)
            if servers:
                target_server = servers[0]
                print(f"Retrieving hosts from {target_server}...")
                
                result = await client.api_query(
                    mgmt_name=target_server,
                    command="show-hosts",
                    details_level="standard"
                )

                if result.success:
                    print(f"Successfully retrieved {result.total} hosts:")
                    for host in result.objects[:5]:
                        print(f"  • {host.get('name')} -> {host.get('ip-address')}")
                else:
                    print(f"API Error: {result.message}")

    finally:
        # 4. Clean up connection pools
        await engine.dispose()

if __name__ == "__main__":
    asyncio.run(main())

3. Using arodonata[mcp] in Projects

When installed with the [mcp] extra, Arodonata can expose Check Point management operations to AI assistants and LLM agents (Claude Code, Claude Desktop, Cursor, Antigravity) via the Model Context Protocol over streamable HTTP.

Pattern A: Standalone Team Server (arodonata-mcp)

Run a dedicated MCP server for your team:

# Install with mcp extra
uv pip install "arodonata[mcp]"

# Launch the streamable-HTTP server
arodonata-mcp --host 0.0.0.0 --port 8765

The CLI loads .env.lib and .env.secrets, resolves credentials via API_KEY_VARS, verifies bearer tokens (ARODONATA_MCP_TOKEN_VARS), and provides DNS-rebinding protection.

Pattern B: Embedded in FastAPI / ASGI Applications (arodonata.mcp)

Embed Arodonata's MCP server into an existing FastAPI or Starlette application, and register custom application-specific tools alongside Check Point tools:

from fastapi import FastAPI
from sqlalchemy.ext.asyncio import create_async_engine
from arodonata import ArodonataClient, ArodonataSettings
from arodonata.mcp import ArodonataMCPSettings, create_asgi_app, create_mcp_server

# Build engine, client, and MCP settings
engine = create_async_engine(database_url)
client = ArodonataClient(engine=engine, settings=settings)
mcp_settings = ArodonataMCPSettings(port=8000, path="/mcp")

# Create the MCP server and add your own custom tools
mcp_server = create_mcp_server(client, mcp_settings, name="firewall-assistant")

@mcp_server.tool()
async def acme_change_ticket(rule_uid: str) -> dict[str, str]:
    """Look up an internal change ticket that introduced a rule."""
    return {"rule_uid": rule_uid, "ticket": "CHG-0001"}

# Create mountable ASGI app
mcp_app = create_asgi_app(client, mcp_settings, server=mcp_server)

# Mount inside your FastAPI app
app = FastAPI(lifespan=lifespan)
app.mount("/", mcp_app)

See examples/09_mcp_embedded.py for the complete runnable example.

---

🏛 Architecture

Arodonata employs a structured Ports and Adapters (Hexagonal) architecture to separate the business models from infrastructure/database implementations.

┌────────────────────────────────────────────────────────────────────────┐
│                        Client Application                              │
│         (manages SQLAlchemy AsyncEngine & environment config)          │
├────────────────────────────────────────────────────────────────────────┤
│                       Arodonata Core Framework                         │
│  ┌───────────────────────┐  ┌───────────────────────┐  ┌─────────────┐ │
│  │       API Layer       │  │       ASDK Core       │  │ Cache Layer │ │
│  │                       │  │                       │  │             │ │
│  │    ArodonataClient    │  │      AMgmtClient      │  │ CacheRepo   │ │
│  │    (High-Level Facade)│  │   (Low-Level Facade)  │  │ DbManager   │ │
│  └───────────┬───────────┘  └───────────┬───────────┘  └──────┬──────┘ │
│              │                          │                     │        │
│              └──────────────────────────┼─────────────────────┘        │
│                                         ▼                              │
│                                 [ ApiPort / CachePort ]                │
├────────────────────────────────────────────────────────────────────────┤
│                         External Infrastructure                        │
│          ┌───────────────────┬───────────────────┬──────────────┐      │
│          │   PostgreSQL DB   │  Check Point API  │  Python env  │      │
│          │  (Cache Storage)  │  (Remote Server)  │ (Pydantic v2)│      │
│          └───────────────────┴───────────────────┴──────────────┘      │
└────────────────────────────────────────────────────────────────────────┘

• API Facade (ArodonataClient): Main developer entry point. Integrates database caching, rate-limiting, and network calls transparently.
• ASDK Core (AMgmtClient): Manages individual sessions, keeps SIDs alive, and maps requests to the official Check Point API.
• Cache Engine (CacheRepository): Manages object serializations, handles TTL checking, and stores entities as JSONB configurations in PostgreSQL.

---

🤖 MCP Server

Arodonata can serve any MCP client (Claude Code, Claude Desktop, Cursor, Antigravity, or your own agents) over streamable HTTP, exposing the Check Point show_* tool surface answered directly from Arodonata's cache with automatic session handling and rate limiting.

Tool Capabilities

• Cached Queries: show_hosts, show_networks, show_groups, show_gateways_and_servers, show_domains, show_object.
• Rulebases: show_access_rulebase, show_nat_rulebase, show_https_rulebase, show_threat_rulebase with formats: markdown (clean table), model_friendly, or raw JSON.
• Native & Sync Operations: arodonata_init (server & domain overview), search_objects (cross-domain search), refresh_objects (smart/incremental cache re-sync), refresh_rulebases, and api_call (read-only by default, gated write mode).
• Live Compatibility: ~40 live tools mirroring @chkp/quantum-management-mcp.
• Declarative CPCRUD (Opt-In): cpcrud_validate, cpcrud_plan, cpcrud_apply (dry-run by default), cpcrud_inverse for reversible Policy-as-Code changes.

Running the Server

# 1. Install with MCP support
uv pip install "arodonata[mcp]"

# 2. Launch the team server
arodonata-mcp --host 0.0.0.0 --port 8765

Client Setup

Claude Code

claude mcp add --transport http arodonata http://127.0.0.1:8765/mcp --header "Authorization: Bearer ${DEMO_MCP_TOKEN}"

Claude Desktop

In your claude_desktop_config.json:

{
  "mcpServers": {
    "arodonata": {
      "url": "http://127.0.0.1:8765/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_BEARER_TOKEN"
      }
    }
  }
}

For full documentation on authentication modes (static vs host), TLS deployment, embedding in your own FastAPI application, and environment variables, see docs/mcp/index.md.

---

📖 Documentation

Complete, searchable documentation is built with MkDocs and served locally or via static hosting:

• Getting Started – Installation, environment variables, and a minimal first script.
• Core Architecture – Ports and adapters, caching/sync, sessions, and multi-domain resolution.
• Configuration Guide – Every ArodonataSettings field and multi-server setup.
• MCP Server – Standalone and embedded streamable-HTTP MCP server, authentication, and the tool list.
• API Reference – Generated reference for every public class and function in arodonata.
• Examples – Runnable, narrated scripts covering queries, search, refresh, and rulebases.

Local Doc Server

Run the local server to explore the documentation offline:

uv run mkdocs serve

Then visit http://localhost:8000 in your web browser.

---

🤝 Contributing

We welcome pull requests! To get set up for local development:

1. Clone the repository:

   git clone https://github.com/chkp-antonr/arodonata
   cd arodonata

2. Install dependencies:

   uv sync --dev

3. Format and check code quality:

   uv run ruff check --fix
   uv run ruff format .
   uv run mypy src/arodonata

4. Run unit tests:

   uv run pytest

---

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

Copyright (c) 2024-2026 Anton Razumov (arazumov@checkpoint.com)