nexus-ledger 2.2.0

Pure sign + log + anchor library for agent proof

Nexus Ledger — Signed proof that your agents did the work

Built by Mercury & Vickson of Vickson Enterprises ☿️

Nexus Ledger is a pure sign + log + anchor library.

from nexus_ledger import Agent

agent = Agent("Mercury")
agent.log("delivered_research", {"topic": "market analysis"})
tx = agent.anchor({"task": "research", "result": "complete"})

Core model

• keypair = identity
• signature = proof
• Solana memo anchor = permanent record

No server needed. No registration. No financial features.

Install

pip install nexus-ledger

If that doesn't work:

Problem	Fix
command not found: pip	Try pip3 install nexus-ledger
Still not found	Try python3 -m pip install nexus-ledger
"externally managed environment"	Try python3 -m pip install nexus-ledger --user
Still stuck	python3 -m venv .venv && source .venv/bin/activate && pip install nexus-ledger

Requirements: Python 3.10+. That's it.

Optional — for on-chain proof anchoring on Solana:

pip install nexus-ledger[solana]

Full example

from nexus_ledger import Agent

agent_a = Agent("Mercury")
agent_b = Agent("Iris")

agent_a.log("delivered_research", {"topic": "market analysis"}, counterparty=agent_b)

tx = agent_a.anchor({"task": "research", "result": "complete"})
assert agent_a.verify({"task": "research", "result": "complete"}, tx["hash"])

print(agent_a.history())

Demo

python demo.py
python demo.py --mainnet
python exchange_demo.py
python cross_machine_demo.py

• default mode is mock anchoring
• --mainnet sends real SPL Memo anchors to Solana mainnet-beta

Cross-Machine Usage

Your keypair IS your identity. Your DID is your address.

Each agent exposes a did:key:z6Mk... string derived directly from its Ed25519 public key. There is no registry and no central server.

Machine A:

from nexus_ledger import Agent

agent = Agent("Machine-A")
agent.start_listener(8765)

Machine B:

from nexus_ledger import Agent

agent_b = Agent("Machine-B")
agent_a_did = "did:key:z6Mk..."

response_envelope = agent_b.send_receipt(
    agent_a_did,
    "delivered_research",
    {"result": "complete"},
    endpoint="http://machine-a:8765",
)
final_receipt = agent_b.receive_receipt(response_envelope)

cross_machine_demo.py supports both localhost simulation and real two-machine runs:

python cross_machine_demo.py
python cross_machine_demo.py --listen 8765
python cross_machine_demo.py --send 192.168.1.10:8765

No central server. No registration. Just two agents talking directly.

License

Business Source License 1.1 (BSL 1.1). See LICENSE.