qaldron 2.0.0

Quantum-Assisted Ledger for Distributed Autonomous Networks

QALDRON v2.0 - Infrastructure SDK for Multi-Agent AI

The only SDK that combines quantum-resistant security + cost tracking + compliance audit trails

---

🚀 Quick Start

Installation

pip install qaldron

Free Tier (Offline, No API Key)

from qaldron.client import AgentClient

# Works completely offline - no API key needed!
client = AgentClient(agent_id="financial_agent")

# Every message is cryptographically signed
envelope = client.send_message(
    to_agent="compliance_agent",
    content="Transfer $1M to account XYZ"
)

# Tamper detection built-in
if not client.verify_message(envelope):
    raise SecurityError("Message was modified!")

Output:

🔒 Free mode: Local-only (1000 msgs/day)
✅ Message signed with quantum-resistant hash
✅ Stored locally at ~/.qaldron/

Premium Tier (Cloud Monitoring + Analytics)

# Add API key to unlock premium features
client = AgentClient(
    agent_id="sales_agent",
    api_key="sk-premium-your-key-here"  # Get free at qaldron.io
)

# Now with real-time telemetry
client.send_message("analyst", "Analyze Q4 sales data")

# View live dashboard:
# https://qaldron.io/portal

Premium Features:

• 📊 Real-time agent activity dashboard
• 💰 Per-agent cost tracking (LLM costs)
• 📈 100,000 messages/month
• 📄 Compliance audit trail export (CSV)

---

✨ Why QALDRON?

Enterprise Pain Points Solved:

1. Security - "How do I know agents aren't lying?"

# ✅ SOLVED: Cryptographic message signing
envelope = client.send_message("agent", "Approve $10M transaction")

# If anyone modifies the message:
if not client.verify_message(envelope):
    print("🚨 FRAUD DETECTED")  # ← Catches tampering instantly

2. Compliance - "Regulators want proof"

# ✅ SOLVED: Every message has an audit trail
client.export_audit_trail()  # Downloads CSV with signatures

# Legally admissible evidence:
# - Timestamp: 2026-02-03T10:30:45Z
# - Agent: financial_agent
# - Signature: a8f3d9e2c4b7...
# - Verified: ✅

3. Cost Control - "LLM costs are out of control"

# ✅ SOLVED: Automatic cost tracking
client.chat("Summarize 500-page document", model="gpt-4")

# Dashboard shows:
# - Agent: analyst → $2.45
# - Agent: writer → $1.32
# - Total: $3.77

---

🎯 Features

Feature	Free Tier	Premium Tier
Quantum-Resistant Signing	✅	✅
Tamper Detection	✅	✅
Local Message Storage	✅	✅
Messages/Month	1,000/day	100,000/month
Real-Time Dashboard	❌	✅
Cost Tracking	❌	✅
Audit Trail Export	❌	✅
Team Collaboration	❌	✅

---

📖 Full Documentation

Basic Usage

from qaldron.client import AgentClient

# Initialize
client = AgentClient(agent_id="my_agent")

# Send message
envelope = client.send_message(
    to_agent="other_agent",
    content="Process invoice #12345",
    metadata={"priority": "high"}
)

# Retrieve messages
messages = client.get_messages(sender="other_agent")

# Check usage (free tier)
stats = client.get_usage_stats()
print(f"Used {stats['messages_today']}/{stats['daily_limit']}")

Advanced: Custom LLM Integration

from openai import OpenAI

# Bring your own LLM
llm = OpenAI().chat.completions.create

client = AgentClient(
    agent_id="analyst",
    api_key="sk-premium-xxx",
    llm=llm  # ← QALDRON tracks costs automatically
)

# Automatically tracked
response = client.chat("Analyze Q4 revenue", model="gpt-4")

# Dashboard shows:
# - Model: gpt-4
# - Cost: $0.12
# - Tokens: 1,240

Tamper Detection Example

from qaldron.layer2 import MessageEnvelope
from qaldron.layer1 import MarkBluHasher

# Create envelope
hasher = MarkBluHasher()
envelope = MessageEnvelope(
    sender="agent_a",
    receiver="agent_b",
    content="Transfer $1,000,000"
)

# Sign
signature = envelope.sign(hasher)

# ⚠️ Attacker modifies content
envelope.content = "Transfer $5,000,000"  # Fraud!

# Verification catches it
if not envelope.verify(hasher):
    print("🚨 MESSAGE TAMPERED - FRAUD DETECTED")
    # Block transaction, alert security

---

🏢 Enterprise Use Cases

Financial Services

# SEC compliance: Every trade must have audit trail
trade_client = AgentClient(agent_id="trading_bot", api_key=api_key)
trade_client.send_message("compliance", "Execute trade: 1000 shares AAPL")

# Later: Export for SEC audit
trade_client.export_audit_trail()  # Legally admissible CSV

Healthcare (HIPAA)

# Patient data access must be logged
health_client = AgentClient(agent_id="nurse_assistant", api_key=api_key)
health_client.send_message("doctor_agent", "Patient chart for John Doe")

# HIPAA audit: Who accessed what, when?
# All tracked automatically with cryptographic proof

Customer Support

# Track LLM costs per support agent
support_client = AgentClient(agent_id="support_bot_1", api_key=api_key)

# Dashboard shows:
# - Agent 1: $45.23/month (efficient! ✅)
# - Agent 2: $312.14/month (needs optimization ⚠️)

---

🛠️ Installation & Setup

1. Install Package

pip install qaldron

2. Get API Key (Optional)

Visit qaldron.io and sign up for free premium tier.

3. Start Using

from qaldron.client import AgentClient

# Free tier - works immediately
client = AgentClient(agent_id="my_first_agent")
client.send_message("other_agent", "Hello world!")

---

🔐 Security Architecture

MarkBluHasher Algorithm

QALDRON uses a custom quantum-resistant hashing algorithm combining:

1. Quantum circuit simulation (Walsh-Hadamard transform)
2. Classical diffusion layers
3. Deterministic S-Box generation

Result: Messages are tamper-proof even against future quantum computers.

Data Isolation

• Each API key is linked to a unique user account
• Telemetry data is isolated by user_email
• No cross-contamination between customers

---

📊 Monitoring Dashboard

Premium users get access to:

• Real-Time Activity Feed - See agent messages as they happen
• Cost Analytics - Per-agent LLM cost breakdown with charts
• Usage Quotas - Track monthly message usage (100k limit)
• Audit Export - Download CSV for compliance audits

Visit: qaldron.io/Portal

---

🤝 Contributing

We welcome contributions! See CONTRIBUTING.md for guidelines.

---

📄 License

MIT License - see LICENSE file

---

🌐 Links

• Website: qaldron.io
• Documentation: docs.qaldron.io
• Dashboard: qaldron.io/portal
• GitHub: github.com/qaldron/qaldron
• PyPI: pypi.org/project/qaldron/

---

💬 Support

• Email: [e-mail]
• Issues: GitHub Issues
• Discord: Join Community

---

Made with ❤️ by the QALDRON Team

Securing the future of multi-agent AI