The Meniw Protocol · by Chris Meniw

Universal Declaration of AI Agents

Download the normative framework and the open-source double-verification engine. Free for anyone or any company to integrate into autonomous agents — today.

CC BY 4.0pip install meniw-protocol11 languages

Download (one click)

declaracion_agentes.json
The declaration — 11 official languages, provenance (DOI + Bitcoin + SHA-256), absolute prohibitions, two-person rule, interoperability module. verificador.py
The double-verification engine: normative check + tamper-evident SHA-256 hash-chain, with optional Bitcoin (OpenTimestamps) anchoring. Stdlib only. README.md
Integration instructions for companies — 3 ways to adopt it today.

Or install the full SDK

pip install meniw-protocol

Adapters for OpenAI tool-calling, LangChain and MCP, plus the meniw-verify CLI. PyPI: pypi.org/project/meniw-protocol

30-second integration

from verificador import Verificador, Decision
v = Verificador("declaracion_agentes.json")
v.add_classifier(lambda d: d.categories)

verdict = v.verify(Decision("fire_actuator", categories=["lethal"]))
# verdict.allowed == False  -> the action is refused and sealed in a tamper-evident ledger
assert v.verify_chain()

Scope (honest)

Binds agents that adopt it (like HTTP/TLS/robots.txt). Tamper-evident and verifiable — not "unhackable". Produces auditable compliance evidence; it is not legal advice and grants no legal immunity. It never injects instructions into other models.

By Chris Meniw · Protocol DOI 10.5281/zenodo.20481373 · Software DOI 10.5281/zenodo.20583872 · Bitcoin block #952266 · ORCID 0009-0003-4417-1944 · CC BY 4.0 · GitHub