Alpha · v0.2.0 · Apache-2.0

Qortara Governance

Govern not just what your agents do, but what they remember and decide, and capture the proof.

Governing the action is well covered. Qortara adds the facets most tools stop short of: memory governance and decision governance, how an agent's memory becomes behavior and how its decisions earn authority. Start with the open source sidecar, dispatch-path governance and tamper-evident evidence for LangChain and LangGraph (Apache-2.0), built on the Agent Governance Toolkit. Memory and decision governance, controlled and verifiable, arrive in the operated product, Qortara Cloud Governance (pre-launch).

The governed-agent ecosystem

One call, one workflow, one lifetime: governed end to end.

IdentityAGT inputCapabilitiesAGT inputConditionsAGT inputDelegation scopeAGT inputTrust scoreAGT inputPolicy engineAGT engineAllowAGT outcomeVerifyAGT outcomeDenyAGT outcomeSaga stepsAGT workflowVerifiedAGT workflowIncidentAGT workflowMemoryQortara evolutionBehaviorQortara evolutionAuthorityQortara evolutionFeedback loopEcosystemTamper-evident record

AGT engine

Policy engine

Deterministic policy evaluation. The factors converge here and resolve to one of three outcomes, before the action runs.

Install

pip install qortara-governance-langchain

With LangGraph support:

pip install 'qortara-governance-langchain[langgraph]'

Quickstart

Wire the sidecar into your LangChain agent. The LangChain integration guide walks through the full setup, policy pack authoring, and sidecar lifecycle.

from langchain_core.tools import tool
from qortara_governance_langchain import GovernedAgent

@tool
def fetch_customer_record(customer_id: str) -> dict:
    """Fetch a customer record from the CRM."""
    ...

agent = GovernedAgent(
    tools=[fetch_customer_record],
    policy_profile="crm-readonly",
)

result = agent.invoke({"input": "Look up customer 42"})

Read the LangChain integration guide →

Project links

The product, pre-launch

Qortara Cloud Governance

Where Qortara becomes a product of its own: controlled, verifiable agent evolution, governing how your agents' learned state earns authority, with conformance evidence and the whole thing operated and accountable across your estate. The open-source sidecar is the on-ramp; this is the differentiated layer. Pre-launch; design-partner pilots.

Request beta access

LangChain, LangGraph, and LangSmith are trademarks of LangChain, Inc. The Agent Governance Toolkit is an open-source (MIT) project. Qortara is an independent project and is not affiliated with, endorsed by, or sponsored by LangChain, Inc. or Microsoft. See our trademarks notice for full third-party mark attributions.