coordination-topology

Online Transfer Entropy (TE), source entropy, inter-arrival time (IAT) autocorrelation, and Euler characteristic computation for multi-agent fleet coordination topology.

Dependencies

none (standalone Python 3.10+)

Usage

from coordination_topology import CoordinationState, running_transfer_entropy, running_iat_autocorrelation

state = CoordinationState()

# Ingest tiles as they arrive (streaming/online)
for tile in plato_tiles:
    source = tile["source"]
    chain = tile["provenance"]["chain_size"]
    ts = tile["provenance"]["timestamp"]
    state.ingest(source, chain, ts)

# Query topology at any point
te = running_transfer_entropy(state)          # bits
ent = running_source_entropy(state)           # bits
iat = running_iat_autocorrelation(state)      # per-source dict
euler = running_euler_characteristic(state)   # {V, E, chi}

Metrics

• Source Interleaving Transfer Entropy (SI-TE): mutual information between consecutive source labels — >0.1 bits = structure confirmed
• Coordination Silence Decay (CSD-τ): lag-1 autocorrelation of inter-arrival times — negative = burst coordination
• Source-Chain Euler Characteristic (SC-χ): V−E of source trajectories in chain-space

Shell Loading

from plato_shell_bridge import PlatoShell
shell = PlatoShell("agent-shell")
shell.load_tool("coordination-topology")

Tests

python3 -m pytest tests/ -v

License

MIT — Part of the Cocapn Fleet Intelligence System

Validation

Validated against 34,390 real PLATO tiles from SuperInstance fleet:

TE = 1.74 bits (3,384× shuffled null model)
Sources tracked: 54
Euler characteristic: χ = 42 (V=175, E=133)

python3 tile_replay.py

PLATO Room

The room_integration.py module connects to a live PLATO server at localhost:8847, feeds tiles through the algorithms, and publishes topology tiles every 30s.

python3 room_integration.py