keeper-pam-webrtc-rs 1.6.2

Keeper PAM WebRTC for Python - A secure, stable, and high-performance Tube API for Python, providing WebRTC-based secure tunneling with enterprise-grade security and reliability optimizations.

Keeper PAM WebRTC for Python

A secure, stable, and high-performance Tube API for Python, providing WebRTC-based secure tunneling with enterprise-grade security and reliability optimizations.

Core Values

Security • Stability • Performance - Built for Keeper Security's mission-critical applications:

• 🔒 Security First: Memory-safe Rust implementation with comprehensive bounds checking
• 🛡️ Enterprise Stability: Lock-free architecture eliminates race conditions and deadlocks
• ⚡ Optimized Performance: Advanced optimizations deliver exceptional speed when you need it
• 🔧 Production Ready: Zero-configuration reliability for demanding security applications

Description

keeper-pam-webrtc-rs provides Python bindings to a Rust-based Tube API for secure communication, designed for:

• Secure tunneling via WebRTC data channels with memory-safe operations
• Multi-connection management through tube abstractions
• Reliable peer connection handling with comprehensive error handling
• Efficient channel management for different communication patterns
• Cross-platform compatibility (Linux, macOS, Windows, Alpine)
• Mission-critical reliability for security-focused applications

This package is designed to be used with Keeper Gateway and Keeper Commander. It provides a secure, reliable tube-based communication system built on WebRTC, specifically tailored for Keeper Security's internal products and security-critical tunneling use cases.

Note: This package is intended for internal Keeper Security products and is not being actively advertised for general use.

Installation

pip install keeper-pam-webrtc-rs

Usage

import keeper_pam_webrtc_rs

# Create a tube registry
registry = keeper_pam_webrtc_rs.PyTubeRegistry()

# Define a signal callback for WebRTC events
def on_signal(signal_dict):
    print(f"Received signal: {signal_dict}")
    # Handle ICE candidates, connection state changes, etc.

# Create a server-side tube for tunneling
server_result = registry.create_tube(
    conversation_id="tunnel-session-123",
    settings={
        "conversationType": "tunnel",
        "target_host": "127.0.0.1", 
        "target_port": "22"  # SSH tunnel example
    },
    trickle_ice=True,
    callback_token="server-token",
    ksm_config="server-config",
    signal_callback=on_signal
)

# Get the offer SDP to send to the client
# NOTE: All SDP (offers, answers) are base64-encoded - use them directly, don't decode!
server_offer = server_result['offer']  # Base64-encoded WebRTC offer
server_tube_id = server_result['tube_id']

# Create a client-side tube with the offer
client_result = registry.create_tube(
    conversation_id="tunnel-client-123",
    settings={
        "conversationType": "tunnel",
        "target_host": "192.168.1.100",
        "target_port": "22"
    },
    trickle_ice=True,
    callback_token="client-token",
    ksm_config="client-config",
    offer=server_offer,  # Pass base64-encoded offer directly (don't decode!)
    signal_callback=on_signal
)

# Get the answer SDP to send back to server
client_answer = client_result['answer']  # Base64-encoded WebRTC answer
client_tube_id = client_result['tube_id']

# Set the remote description on the server
# NOTE: Pass base64-encoded answer directly
registry.set_remote_description(server_tube_id, client_answer, is_answer=True)

# Check connection state
state = registry.get_connection_state(server_tube_id)
print(f"Connection state: {state}")

# Close when done
registry.close_tube(server_tube_id)
registry.close_tube(client_tube_id)

Server Mode with TCP Listener

For server tubes that listen for external TCP connections:

# Create server tube with TCP listener (dynamic port)
server_result = registry.create_tube(
    conversation_id="tcp-tunnel",
    settings={
        "conversationType": "tunnel",
        "local_listen_addr": "127.0.0.1:0"  # 0 = dynamic port assignment
    },
    trickle_ice=True,
    callback_token="token",
    ksm_config="config",
    signal_callback=on_signal
)

# Get actual listening address (port assigned by OS)
listen_addr = server_result['actual_local_listen_addr']  # "127.0.0.1:59194"
host, port = listen_addr.split(':')

# External clients can now connect to this address
import socket
client = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
client.connect((host, int(port)))
# Data flows: TCP → WebRTC → Remote tube → Remote target

API Documentation: See docs/PYTHON_API_CONTRACT.md for complete API reference including:

• Base64 SDP encoding requirements
• Return value specifications
• Migration guides and common pitfalls

Features

• 🔒 Memory Safety: Rust-powered implementation prevents buffer overflows and memory corruption
• 🛡️ Reliable Architecture: Lock-free design eliminates race conditions and ensures stability
• ⚡ Efficient Performance: Optimized for speed without compromising security or stability
• 🌊 Tube Abstraction: High-level API for managing WebRTC-based secure tunnels
• 🌍 Cross-Platform: Secure, consistent behavior across Linux, macOS, Windows, Alpine
• 🐍 Python Integration: Built with abi3 for maximum compatibility (Python 3.7+)
• 🔧 Production Hardened: Comprehensive error handling and graceful degradation

Tube API Architecture

This implementation provides a Tube-based abstraction over WebRTC:

Security Features

• Memory-Safe Operations: Rust's ownership system prevents common security vulnerabilities
• Bounds Checking: Comprehensive validation prevents buffer overflows and data corruption
• Zero Unsafe Code: Hot paths use only verified, safe Rust code (except vetted SIMD intrinsics)
• Graceful Error Handling: Robust error recovery prevents crashes and data leaks

Tube Management

• Multi-Connection Support: Each tube can manage multiple WebRTC connections
• Channel Abstraction: High-level channel management for different protocols
• State Management: Comprehensive connection state tracking and reporting
• Signal Handling: Event-driven architecture for ICE candidates and state changes

Performance Features

• SIMD Optimization: Hardware-accelerated frame parsing with safe fallbacks
• Zero-Copy Pipelines: Efficient data handling minimizes memory overhead
• Event-Driven Design: Native WebRTC events provide responsive communication
• Always Optimized: Maximum efficiency by default, no configuration required

Tube API Reference

Core Methods

• create_tube(conversation_id, settings, ...) - Create a new secure tube or add conversation to existing tube
• set_remote_description(tube_id, sdp, is_answer) - Set remote SDP description
• add_ice_candidate(tube_id, candidate) - Add ICE candidate for connection
• get_connection_state(tube_id) - Get current connection state
• close_connection(connection_id) - Close specific connection
• close_tube(tube_id) - Close entire tube

Conversation Types

The tube API supports different communication patterns:

• tunnel - Secure TCP tunneling through WebRTC
• guacd - Apache Guacamole protocol tunneling
• socks5 - SOCKS5 proxy tunneling

Build & Verification

To build and verify the implementation:

# Standard build (all optimizations enabled)
cargo build --release

# Run comprehensive test suite
cargo test --release

Why This Implementation?

Built specifically for Keeper Security's tunneling requirements:

• Security-First Design: Memory safety and comprehensive validation prevent vulnerabilities
• Mission-Critical Reliability: Lock-free architecture ensures stable operation under load
• Optimized for Security Applications: Performance optimizations that don't compromise security
• Tube Abstraction: High-level API designed specifically for secure tunneling use cases

The secure, stable, high-performance tube communication system for enterprise security applications.