Skip to main content

Post-quantum AI decision attestation for AutoGen, LlamaIndex, and any Python AI pipeline

Project description

autogen-rubric

Post-quantum AI decision attestation for AutoGen, LlamaIndex, and any Python AI pipeline.

Every decision your AI agents make, cryptographically signed and anchored permanently to Hedera Consensus Service. Built for EU AI Act Article 12 compliance. Zero required dependencies.

pip install autogen-rubric

Quick Start

from autogen_rubric import RubricClient

client = RubricClient(api_key="your-rubric-api-key")

result = client.attest(
    agent_id="loan-underwriting-agent-v2",
    output="Application #12345 approved. Confidence: 0.94.",
    confidence=0.94,
)

print(result.attestation_id)  # Verifiable forever on Hedera
print(result.stage)           # "pending" -> "anchored"

Get a free API key at rubric-protocol.com.

AutoGen Integration

from autogen import AssistantAgent, UserProxyAgent
from autogen_rubric import RubricClient, RubricAutoGenHook

client = RubricClient(api_key="your-key", background_queue=True)
hook = RubricAutoGenHook(client=client, pipeline_id="compliance-pipeline-v1")

assistant = AssistantAgent("assistant", llm_config=llm_config)
hook.register(assistant)  # Every message now attested

user_proxy = UserProxyAgent("user")
user_proxy.initiate_chat(assistant, message="Review this loan application...")

LlamaIndex Integration

from llama_index.core.callbacks import CallbackManager
from autogen_rubric import RubricClient, RubricLlamaIndexHandler

client = RubricClient(api_key="your-key", background_queue=True)
handler = RubricLlamaIndexHandler(client=client)
callback_manager = CallbackManager([handler])

Function Decorator

from autogen_rubric import RubricClient, attest_function

client = RubricClient(api_key="your-key")

@attest_function(client=client, agent_id="risk-scorer-v2")
def score_risk(application: dict) -> str:
    return "approved" if application["score"] > 0.7 else "declined"

Configuration

Parameter Default Description
api_key required Your Rubric API key
node "us" Federation node: us, sg, jp, ca, eu, auto
enterprise False Merkle batching for high volume
background_queue False Non-blocking attestation
flush_interval 5.0 Seconds between queue flushes
timeout 15 HTTP timeout in seconds

Requirements

  • Python 3.9+
  • Zero required dependencies for core
  • AutoGen: pip install autogen-rubric[autogen]
  • LlamaIndex: pip install autogen-rubric[llama]
  • Both: pip install autogen-rubric[all]

EU AI Act Article 12

Enforcement deadline: August 2, 2026.

rubric-protocol.com | Scott@Rubric-Protocol.com Echelon Intelligence Group LLC | Wyoming, USA

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

autogen_rubric-1.3.0.tar.gz (7.6 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

autogen_rubric-1.3.0-py3-none-any.whl (7.0 kB view details)

Uploaded Python 3

File details

Details for the file autogen_rubric-1.3.0.tar.gz.

File metadata

  • Download URL: autogen_rubric-1.3.0.tar.gz
  • Upload date:
  • Size: 7.6 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.10.12

File hashes

Hashes for autogen_rubric-1.3.0.tar.gz
Algorithm Hash digest
SHA256 2233bd6a52636200edfe3f90b77d6c1e1245851223ff63ed76947d01e14c19e4
MD5 6fc445eb13e477eb56fe466f73cb8d1b
BLAKE2b-256 659866b63bddcbb732e1081374ee46bc20cd3d2af09ca629b6aad92f5df51bc6

See more details on using hashes here.

File details

Details for the file autogen_rubric-1.3.0-py3-none-any.whl.

File metadata

  • Download URL: autogen_rubric-1.3.0-py3-none-any.whl
  • Upload date:
  • Size: 7.0 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.10.12

File hashes

Hashes for autogen_rubric-1.3.0-py3-none-any.whl
Algorithm Hash digest
SHA256 6acb162e34ef3afb8822e2124380b8a7761ab660ed5c9fb4f91d3de674af9292
MD5 fa020af3800ec9ce0380b75e0c8b1951
BLAKE2b-256 ce32f94eb1981314dcef305226e5b60af4d8b0a3fe306eeb7dc8c7a78bacf2a5

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page