# Best AI Agent Governance (bestaiagentgovernance.com) > Evidence-gated runtime control catalog for AI agents covering MCP gateways, runtime security, non-human agent identity, and observability. ## Control Layers - [CP-01 MCP & AI Gateways](https://bestaiagentgovernance.com/mcp-gateways/): Inline proxies mediating model calls and Model Context Protocol tool invocations with authentication, quotas, and policy hooks. - [CP-02 Agent Security & Runtime Protection](https://bestaiagentgovernance.com/agent-security/): Discovery, shadow-agent inventory, prompt-injection defense, and inline runtime guardrails for autonomous workflows. - [CP-03 Agent Identity & Access](https://bestaiagentgovernance.com/agent-identity/): Non-human credential brokers, delegated OAuth/MCP scopes, conditional access, and revocable token vaults for agents. - [CP-04 Observability, Registry & Audit](https://bestaiagentgovernance.com/observability-audit/): Control planes for registering enterprise agents, monitoring SaaS and workforce agent usage, and logging action trails. - [Controls Ledger](https://bestaiagentgovernance.com/controls/): Verified vendor ledger. - [Method & Disclosure](https://bestaiagentgovernance.com/method/): Two-judge evidence gates and sponsored rules. ## Verified Vendors (15) - Okta (Okta & Auth0 for AI Agents) (agent-identity, 51 cited facts, latest 2026-11-21): https://www.okta.com - Kong AI Gateway (mcp-gateways, 44 cited facts, latest 2026-11-21): https://konghq.com - Portkey (Prisma AIRS AI Gateway) (mcp-gateways, 41 cited facts, latest 2026-11-21): https://portkey.ai - TrueFoundry (mcp-gateways, 38 cited facts, latest 2026-11-21): https://www.truefoundry.com - Straiker (agent-security, 37 cited facts, latest 2026-11-21): https://www.straiker.ai - LiteLLM (mcp-gateways, 36 cited facts, latest 2026-11-21): https://www.litellm.ai - Lasso Security (agent-security, 35 cited facts, latest 2026-11-21): https://www.lasso.security - Obsidian Security (observability-audit, 33 cited facts, latest 2026-11-21): https://www.obsidiansecurity.com - WitnessAI (observability-audit, 33 cited facts, latest 2026-11-21): https://witness.ai - Zenity (agent-security, 27 cited facts, latest 2026-11-21): https://zenity.io - Cloudflare AI Gateway (mcp-gateways, 24 cited facts, latest 2026-11-21): https://developers.cloudflare.com/ai-gateway/ - Cisco AI Defense (agent-security, 23 cited facts, latest 2026-11-21): https://securitydocs.cisco.com/docs/scc/admin/169999.dita - Docker MCP Gateway (mcp-gateways, 23 cited facts, latest 2026-11-21): https://docs.docker.com/ai/mcp-catalog-and-toolkit/mcp-gateway/ - Microsoft Agent 365 (observability-audit, 21 cited facts, latest 2026-11-21): https://learn.microsoft.com/en-us/microsoft-agent-365/ - Palo Alto Networks Idira (formerly CyberArk) (agent-identity, 21 cited facts, latest 2026-11-21): https://www.paloaltonetworks.com/idira