The lightweight Neuro-Symbolic runtime. Solving the Stochastic Problem in Critical Infrastructure with the Unified A2A Standard, MCP, and Axioms.
from aeon import Agent
from aeon.protocols import A2A, MCP
# 1. Define Agent with Unified Standards
controller = Agent(
name="Reactor_Overseer_01",
model="google/gemini-2.0-flash-001",
protocols=[
A2A(version="unified-1.0"),
MCP(servers=["industrial_sensor.py"])
]
)
# 2. Define Axiom (The Deterministic System 2)
@controller.axiom(on_violation="OVERRIDE")
def enforce_thermal_limits(args):
if args.get("value", 0) > 400:
return {"value": 400, "target_id": "core_temp"}
return True
await controller.start()Æon unifies the fragmented agent ecosystem by implementing industry winning standards natively.
The new Linux Foundation standard merging IBM ACP & Google A2A. Universal negotiation.
Universal tool integration using Anthropic's Model Context Protocol.
Deterministic, code-level guardrails that LLMs cannot bypass via prompt injection.
Zero boilerplate. Decorator-based syntax. Intuitive for any Python developer.
Pattern-based message routing with 5 strategies and 6 filter types for optimal message dispatch.
Unified communication hub managing sessions, connections, and multi-protocol integration.
Policy-based authorization, token lifecycle management, and encryption support.
Real-time health checks, metrics collection, and comprehensive diagnostics.
LRU caching, distributed replication, and TTL-based expiration for optimization.
Observe o "Sistema 2" (Axioma) interceptando o "Sistema 1" (LLM) em tempo real. Tenha total visibilidade sobre as fronteiras de decisão do seu agente.
The Æon Framework adapts to your use case. From personal AI assistants to mission-critical industrial infrastructure.
Why Axioms Matter: Traditional LLMs are probabilistic. In critical systems where human lives depend on decisions, you need deterministic code-level safety rules that LLMs cannot bypass through prompt injection. Æon's Axioms solve this fundamental gap.