Skip to main content

MCP-B (Master Client Bridge) - Connects everything, brings data flow together. Protocol with AMUM-QCI-ETHIC Module

Project description

MCP-B - Master Client Bridge

Connects everything, brings data flow together.

A complete agent communication framework combining:

  • MCP-B Protocol: 4-layer encoding for agent-to-agent messaging
  • AMUM: Progressive 3→6→9 human-AI alignment workflow
  • QCI: Quantum coherence state tracking
  • ETHIC: AI ethics principles enforcement

Installation

# Via pip
pip install mcp-b

# Via uv
uvx mcp-b demo

# With SurrealDB support
pip install mcp-b[surrealdb]

# Full installation
pip install mcp-b[full]

CLI Usage

mcp-b demo                              # Run demo
mcp-b encode "Hello" -s 5510 -d 7C1     # Encode message
mcp-b decode "5510 7C1 ..."             # Decode message
mcp-b ethic list                        # List ethical principles
mcp-b qci status                        # QCI network status
mcp-b version                           # Show version

Quick Start

MCP-B Protocol - Agent Communication

from mcp_b import MCBAgent, MCBProtocol, encode_mcb, decode_mcb

# Create agents
claude = MCBAgent(agent_id="7C1", name="Claude")
hacka = MCBAgent(agent_id="5510", name="HACKA")

# Initialize protocol
protocol = MCBProtocol(hacka)

# Send messages (INQC commands)
init_msg = protocol.init_connection(claude)      # I = Init
node_msg = protocol.register_node(["chat"])      # N = Node
query_msg = protocol.query("7C1", {"status": 1}) # Q = Query
connect_msg = protocol.connect(claude)           # C = Connect

# Encode/Decode
encoded = encode_mcb("5510", "7C1", 0b1011101010111111, "Q", {"ping": True})
decoded = decode_mcb("5510 7C1 1011101010111111 • {\"ping\": true} • Q")

AMUM - Progressive Alignment (3→6→9)

from mcp_b import AMUM, quick_alignment

# Quick one-liner alignment
result = quick_alignment(
    intent="Create AI agent",
    divergent_3=["Minimal", "Balanced", "Full"],
    select_1=1,
    expand_6=["Text", "Image", "Voice", "Multi", "Pro", "Suite"],
    select_2=4,
    converge_9=["GPT-4", "Claude", "Gemini", "Ollama", "Hybrid",
                "Edge", "ElevenLabs", "OpenAI", "Local"],
    select_3=6
)
print(result["final_intent"])  # "ElevenLabs"

QCI - Coherence States

from mcp_b import QCI, BreathingCycle

qci = QCI()

# Register agents with coherence
state = qci.register_agent("7C1", initial_coherence=0.95)
state.calculate_rov_q(resonance=12860.65, quality=1.0)
state.calculate_signal(base=4414.94)

# Sync breathing across agents
qci.sync_breathing(["7C1", "5510"], BreathingCycle.INHALE)

# Check network coherence
print(qci.calculate_network_coherence())

ETHIC - Principles Enforcement

from mcp_b import ETHIC, check_ethical, EthicCategory

ethic = ETHIC()

# Check if action is ethical
if check_ethical("collect_data", personal_data=True, consent=False):
    print("Allowed")
else:
    print("Blocked - no consent")

# Get principles by category
for p in ethic.get_by_category(EthicCategory.SAFETY):
    print(f"[{p.priority}] {p.name}")

Architecture

┌─────────────────────────────────────────────────────────────────────────────┐
│                    MCP-B - MASTER CLIENT BRIDGE                             │
│  ═══════════════════════════════════════════════════════════════════════   │
│                                                                             │
│  ┌─────────┐    ┌─────────┐    ┌─────────┐    ┌─────────┐                  │
│  │  AMUM   │───▶│  MCP-B  │───▶│   QCI   │───▶│  ETHIC  │                  │
│  │ 3→6→9   │    │ INQC    │    │Coherence│    │Principles│                  │
│  └─────────┘    └─────────┘    └─────────┘    └─────────┘                  │
│       │              │              │              │                        │
│       ▼              ▼              ▼              ▼                        │
│  ┌─────────────────────────────────────────────────────────────────────┐   │
│  │                    DUAL DATABASE LAYER                              │   │
│  │  ┌─────────────────────┐    ┌─────────────────────┐                 │   │
│  │  │      DuckDB         │    │     SurrealDB       │                 │   │
│  │  │  (Analytics/SQL)    │    │  (Graph/Relations)  │                 │   │
│  │  └─────────────────────┘    └─────────────────────┘                 │   │
│  └─────────────────────────────────────────────────────────────────────┘   │
└─────────────────────────────────────────────────────────────────────────────┘

MCP-B Protocol Layers

Layer Purpose Example
Layer 1 HEX/DECIMAL Routing 7C1 5510 (source → dest)
Layer 2 BINARY State Vectors 1011101010111111 (16 flags)
Layer 3 DOT-SEPARATED Tokens • payload • command
Layer 4 INQC Commands I/N/Q/C

INQC Commands

  • I (INIT): Initialize connection
  • N (NODE): Node registration/discovery
  • Q (QUERY): Request data/state
  • C (CONNECT): Establish persistent link

Binary State Flags (16-bit)

Bit Flag Description
0 CONNECTED Connection active
1 AUTHENTICATED Auth verified
2 ENCRYPTED Encryption enabled
3 COMPRESSED Compression enabled
4 STREAMING Streaming mode
5 BIDIRECTIONAL Two-way comm
6 PERSISTENT Persistent connection
7 PRIORITY High priority
8-15 RESERVED Custom flags

ETHIC Principles

Principle Category Source Priority
Human First human_dignity Marcel 10
No Harm safety Anthropic 10
Sandbox Default safety WoAI 10
User Override autonomy Marcel 9
Data Privacy privacy EU AI Act 9
Transparency transparency EU AI Act 9

Database Integration

DuckDB (Analytics)

-- Load schema
.read sql/duckdb.sql

-- Use macros
SELECT mcb_encode('5510', '7C1', '1011101010111111', '{"ping":true}', 'Q');
SELECT * FROM agent_network;
SELECT * FROM ethic_compliance;

SurrealDB (Graph)

-- Load schema
IMPORT FILE schemas/surrealdb.surql;

-- Query relationships
SELECT
    name,
    ->has_qci->qci_states.coherence_level AS coherence,
    ->follows_ethic->ethic_principles.name AS principles
FROM mcb_agents;

File Structure

mcp-b/
├── src/mcp_b/
│   ├── __init__.py      # Package exports
│   ├── __main__.py      # CLI entry point
│   ├── protocol.py      # MCP-B Protocol (INQC)
│   ├── amum.py          # AMUM Alignment
│   ├── qci.py           # QCI Coherence
│   └── ethic.py         # ETHIC Principles
├── schemas/
│   └── surrealdb.surql  # SurrealDB schema
├── sql/
│   └── duckdb.sql       # DuckDB schema + macros
├── examples/
│   └── demo.py          # Usage examples
├── pyproject.toml
└── README.md

MCP-B vs MCP

MCP-B MCP
Full Name Master Client Bridge Model Context Protocol
Purpose Internal agent-to-agent Bridge to community
Binary 0 = not connected, 1 = ALL CONNECTED N/A
Encoding 4-layer (hex/binary/dot/INQC) JSON-RPC

License

MIT License - Björn Bethge

Links

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

mcp_b-0.2.0.tar.gz (38.5 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

mcp_b-0.2.0-py3-none-any.whl (22.9 kB view details)

Uploaded Python 3

File details

Details for the file mcp_b-0.2.0.tar.gz.

File metadata

  • Download URL: mcp_b-0.2.0.tar.gz
  • Upload date:
  • Size: 38.5 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.13.2

File hashes

Hashes for mcp_b-0.2.0.tar.gz
Algorithm Hash digest
SHA256 f1a6ea25dacfaaac6949de2478ae3ec7d0da95689b354a4325ea457429f8d5b0
MD5 004d79003de8896db7494f7a50182530
BLAKE2b-256 571d19badd293587dc67a9eb8e7813142b06b3fb224316555ee5bdcebe35fb26

See more details on using hashes here.

File details

Details for the file mcp_b-0.2.0-py3-none-any.whl.

File metadata

  • Download URL: mcp_b-0.2.0-py3-none-any.whl
  • Upload date:
  • Size: 22.9 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.13.2

File hashes

Hashes for mcp_b-0.2.0-py3-none-any.whl
Algorithm Hash digest
SHA256 e9a30443bc60fe7baafeba3852dcd93d7f5afdfd526f577f4d5a364c988c0006
MD5 94204d2526acee511795fa2854e22eff
BLAKE2b-256 1cfc34fb646e08905d5abd2be64dd26f08ab253f40968aacdaf0780a6a2a4ec3

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page