hivemind-bus-client 0.9.0a1

Hivemind Websocket Client

HiveMind Bus Client

Python client library for hivemind-core — the primary library for building HiveMind satellites.

Extends the standard OVOS bus client with secure, encrypted, routed communication across a HiveMind mesh network. See the whitepaper for protocol details.

Install

uv pip install hivemind_bus_client

Quick Start

from hivemind_bus_client import HiveMessageBusClient
from hivemind_bus_client.message import HiveMessage, HiveMessageType
from ovos_bus_client.message import Message

client = HiveMessageBusClient(host="ws://localhost", port=5678)
client.connect()

client.on_mycroft("speak", lambda msg: print(msg.data["utterance"]))

client.emit(HiveMessage(HiveMessageType.BUS,
                        Message("recognizer_loop:utterance",
                                {"utterances": ["hello world"]})))

Message Types

Type	Direction	Description
BUS	satellite ↔ master	Standard AI bus message
ESCALATE	upstream	Forward up the authority chain
QUERY	upstream + response	First answering node wins
BROADCAST	downstream	Admin pushes to all satellites
PROPAGATE	flood	Forward to all peers in all directions
CASCADE	flood + responses	Collect answers from all nodes
INTERCOM	any → any	End-to-end hybrid-encrypted (AES-GCM + RSA)
PING	inside PROPAGATE	Flood-based network topology discovery

Security

• Per-link encryption: AES-GCM or ChaCha20-Poly1305, negotiated at handshake
• Hybrid INTERCOM encryption: Random AES-256 key per message, RSA-encrypted key exchange
• Trusted peers: NodeIdentity.trusted_keys gates BUS injection for PROPAGATE and INTERCOM from untrusted peers
• Topology-as-policy: Network structure determines access control (see whitepaper §4)

Trusted Keys

from hivemind_bus_client.identity import NodeIdentity

identity = NodeIdentity()
identity.add_trusted_key("living-room-hub", "<RSA_PUBLIC_KEY>")
identity.save()

Untrusted PROPAGATE(BUS) and INTERCOM messages are silently dropped.

CASCADE Response Aggregation

CASCADE collects responses from all reachable nodes via CascadeAggregator:

proto.cascade_select_callback = lambda responses: responses[0]  # pick best
proto.hive_mapper = mapper  # enables early resolution when all nodes answered

CLI

hivemind-client set-identity --key KEY --password PASS --siteid SITE
hivemind-client terminal --host ws://hub.local --port 5678
hivemind-client ping --host ws://hub.local --port 5678
hivemind-client send-mycroft --msg "recognizer_loop:utterance" --payload '{"utterances": ["hello"]}'
hivemind-client escalate --msg "recognizer_loop:utterance" --payload '{"utterances": ["hello"]}'
hivemind-client propagate --msg "recognizer_loop:utterance" --payload '{"utterances": ["hello"]}'

Documentation

Full docs in /docs:

• API Reference — HiveMessage, HiveMessageBusClient, NodeIdentity, HiveMapper
• Message Types — Routing modes, QUERY, CASCADE, PING
• Identity & Trust — Credentials, RSA keys, trusted peers
• Examples — Chat, TTS, INTERCOM, QUERY, CASCADE, trust management
• CLI Guide — Terminal, ping, send commands
• Serialization — Binary wire format