---
id: agent-governance-toolkit-core
version: "5.0.0"
license: MIT
license_treatment: permissive
maintenance: active
---

# agent-governance-toolkit-core — Core runtime, kernel, and trust layer for the Agent Governance Toolkit

License: permissive · Maintenance: active · Downloads: 102.0K/mo

## What it is and what it does

agent-governance-toolkit-core consolidates five previously separate packages into a single distribution. It provides core runtime infrastructure for multi-agent systems, including kernel architecture, trust exchange mechanisms, execution supervision with privilege rings, saga orchestration, audit trails, identity management, and policy enforcement for cloud-native agent ecosystems.

The package maintains backward compatibility with the five legacy packages through deprecated redirects. It depends on cryptography, pydantic, jsonschema, and structlog to enforce type safety, validation, and observability. The toolkit is designed for scenarios where agents need to operate under governance constraints with verifiable audit trails and zero-trust security models.

Use it for:

- Building multi-agent systems where audit trails and compliance reporting are mandatory for regulatory or security reasons.
- Implementing privilege isolation and execution rings to prevent unauthorized cross-agent access or privilege escalation.
- Managing agent identity and trust relationships in cloud-native distributed agent deployments.
- Orchestrating complex workflows using saga patterns with automatic compensation and rollback on failure.
- Enforcing policy-driven governance and access control across heterogeneous agent ecosystems.

## Worth the install?

AI-flagged interpretation of the facts above — verify before relying.

Provides a unified runtime kernel and trust layer for building, executing, and governing multi-agent systems with audit trails, privilege isolation, and identity management.

Yes, if you are building multi-agent systems that require governance, audit, or compliance infrastructure. The package is actively maintained, has no known vulnerabilities, and consolidates significant functionality into a single install. Migration from legacy packages is straightforward with backward-compatible redirects provided.

## Install

pip install agent-governance-toolkit-core
uv add agent-governance-toolkit-core
poetry add agent-governance-toolkit-core

## Installing agent-governance-toolkit-core

Before you install: Low friction installation with a pure-Python wheel. Actively maintained with a recent release (11 days old) and substantial community engagement (5929 stars). Requires Python 3.11 or later.

License in practice: MIT license permits unrestricted use, modification, and distribution in both open and proprietary projects with minimal restrictions.

Quickstart:

pip install agent-governance-toolkit-core

import agent_governance_toolkit_core
# See migration guide for import paths from consolidated packages

Requires Python 3.11 or later.

Verify before relying:

- Specific import paths and API entry points for kernel, trust, and governance functionality.
- What optional extras (full, iatp, observability) provide and how to use them.
- Performance characteristics and scalability limits for large multi-agent deployments.
- Specific compliance frameworks the governance layer is designed to support.

## Package facts

- License: MIT (permissive)
- Python support: supports_current
- Install friction: low
- Maintenance: active
- Downloads: 102.0K/month (top 15,000 on PyPI)
- Known vulnerabilities: none known

## Tags

multi-agent runtime governance, agent kernel architecture, trust layer for ai agents, agent audit and compliance, zero-trust agent execution, agent privilege isolation, agent identity management, multi-agent-systems, governance-compliance, audit-trails

[View on SkillFed](https://skillfed.io/packages/agent-governance-toolkit-core) · [View on PyPI](https://pypi.org/project/agent-governance-toolkit-core/)