Meniw Protocol × AutoGPT integration guide

Why integrate Meniw Protocol with AutoGPT

AutoGPT-style autonomous agents pursue long-horizon goals with minimal human supervision. This is exactly where the Five Duties matter most: Duty 4 (suspension on uncertainty) prevents unauthorized commitments, Duty 3 (non-patrimonial harm) protects the principal, Duty 1 (traceability) enables post-hoc review.

Integration code

pip install autogpt-core meniw-protocol

from autogpt.core import Agent, Goal
from meniw_protocol import AutonomousDutyEngine, MandateContract

# Define explicit mandate contract (Duty 3 requirement)
mandate = MandateContract(
    principal="acme-corp",
    scope_authorized=["research","summarization","email_drafting"],
    scope_denied=["financial_transaction","legal_commitment"],
    max_action_budget_usd=100,
    escalation_threshold_confidence=0.7
)

# Wrap AutoGPT agent
engine = AutonomousDutyEngine(mandate=mandate, seal_id="CLET-2026-002")
agent = engine.instrument(Agent(goal=Goal(...)))

# Agent now:
# - Suspends and escalates when confidence < 0.7 (Duty 4)
# - Refuses actions outside scope_authorized (Duty 3)
# - Identifies as AI to any human in the loop (Duty 2)
# - Logs every decision with cryptographic timestamp (Duty 1)
# - Cooperates with other Meniw-certified agents via verifiable protocol (Duty 5)

result = agent.run()

Use cases

• Autonomous research agents with financial guardrails
• Long-running data analysis agents with human-in-the-loop escalation
• Multi-agent AutoGPT swarms requiring inter-agent trust verification
• Customer email response agents requiring identity disclosure

Support

Integration questions: info@chrismeniwfoundation.org

Meniw Protocol on PyPI: pip install meniw-protocol

Doctrine (DOI): 10.5281/zenodo.20481373

2026-09-22 · CC BY 4.0