AGNTCY provides complete open-source infrastructure for AI agent collaboration—discovery, identity, messaging, and observability that works across any vendor or framework. Launched by Cisco in March 2025 and transitioned to Linux Foundation governance in July 2025, AGNTCY enables secure, interoperable multi-agent systems with support for A2A and MCP protocols, backed by over 75 member organizations.
AGNTCY.org
Connect with us on GitHub
EXPERTISE
EXPERTISE
SHARE
Open Infrastructure for the Internet of Agents
-
Open Agent Schema Framework (OASF) enables agents to discover and understand each other’s capabilities across different vendors and frameworks, forming a distributed Internet of Agents inventory.
-
Cryptographically verifiable identity and access control system using decentralized technologies ensures secure, trustworthy cross-organizational agent interactions with robust authentication and authorization.
-
Secure Low-latency Interactive Messaging protocol defines standards for efficient network-level communication between AI agents with MLS encryption and quantum-safe security for multi-modal interactions.
-
OpenTelemetry-native telemetry collectors and evaluation tools enable comprehensive debugging and performance monitoring of multi-agent workflows across distributed systems at enterprise scale.
-
Native support for Agent2Agent (A2A) and Model Context Protocol (MCP) enables dynamic multi-agent environments where agents and MCP servers are discoverable through unified directories.
-
Neutral, community-driven governance under Linux Foundation ensures open standards with backing from Cisco, Dell Technologies, Google Cloud, Oracle, Red Hat, and 75+ supporting organizations.
-
Validated in real-world use cases including AI-driven CI/CD pipelines, telecom network automation at Swisscom, and KPMG’s Clara AI auditing platform for enterprise-scale implementations.
-
Fully open-source infrastructure distributed under Apache 2.0 license with active GitHub repositories, comprehensive documentation, and collaborative community development model.
-
Peer-to-peer directory service allows organizations to run their own directories and keep them synchronized, enabling publication, exchange, and discovery of agent capabilities across the ecosystem.
-
Designed to support growing numbers of agents without performance degradation, enabling sophisticated AI solutions for financial services, healthcare, security compliance, and IT operations.
Expertise Gallery Labels
Demos, docs and playgrounds
Helpful resources
-
Complete documentation covering all core components including OASF, Agent Directory, SLIM messaging, Identity system, Observability tools, and CSIT testing framework.
-
Official project website with mission statement, infrastructure overview, supported technologies, community resources, and links to GitHub repositories and documentation.
-
Official GitHub organization hosting all AGNTCY repositories including ACP SDK, Directory service, Identity system, documentation, website code, and contributing guidelines.
-
Python SDK for Agent Connect Protocol facilitating cross-framework agent interoperability with support for uv and poetry dependency management.
-
Official Python package for Agent Connect Protocol SDK distributed via PyPI, latest version 1.5.2, built with GitHub actions and Apache 2.0 licensed.
-
Distributed announce and discovery service for Multi-Agentic Systems enabling peer-to-peer publication and exchange of agent information with Python SDK.
-
Official Linux Foundation press release announcing AGNTCY project adoption with formative members Cisco, Dell Technologies, Google Cloud, Oracle, and Red Hat.
-
CoffeeAGNTCY reference implementation walkthrough with component-level guides for SLIM messaging, Identity integration, and Agent Directory configuration.
-
Open Agent Schema Framework documentation including extensible data model, server implementation, SDK tools for decoding and validation, and contribution guidelines.
-
Secure Low-latency Interactive Messaging protocol documentation covering core concepts, configuration, authentication, session management, group communication, and integrations.
-
Complete guide for agent discovery and announcement including architecture, trust model, event streaming, CLI/SDK references, and API documentation.
-
Identity management system guide covering agent verification, credentials, identity provider connectivity, policy creation, and API access for Agents and MCP servers.
-
Technical article exploring AI agent interoperability challenges and how AGNTCY enables framework-agnostic multi-agent system development.
-
Comprehensive overview of AGNTCY’s mission to build open standards enabling AI agents from different vendors and platforms to communicate efficiently.
-
Analysis of AGNTCY’s role in solving AI agent fragmentation under Linux Foundation governance with insights on production deployments.
-
Coverage of Cisco’s donation of AGNTCY agentic AI platform to Linux Foundation with technical analysis of infrastructure components.
-
HPC Wire coverage of AGNTCY project launch focusing on standardizing open multi-agent system infrastructure for enterprise deployments.
-
Documentation repository with contributing guidelines, governance information, and code of conduct for participating in AGNTCY open source development.
-
Core maintainer Galileo’s perspective on building the future of multi-agentic systems through industry collaboration and open standards.
-
Industry research analysis of AGNTCY’s transition to Linux Foundation governance and implications for enterprise AI agent deployment strategies.
