Kong Context Mesh is a new product for automatic enterprise API discovery. It converts them into agent tools and ensures rollout with runtime governance. The solution is designed to assist in the transition from API-first to agent-first architectures.
According to Marco Palladino, CTO and co-founder of Kong, organizations can reuse their existing API investments. “Organizations have spent years building APIs as the nervous system of the enterprise. Context Mesh allows them to reuse that investment to power agents instead of starting from scratch,” said Palladino. He emphasizes that the goal is to help customers move from experimenting with agents to reliable operations at scale.
Context Mesh leverages knowledge that the Kong Konnect platform already manages about enterprise environments, including endpoints, schemas, authentication requirements, and policies. The system provides automatic detection of all APIs under management, capabilities to select precise context for specific agents, and instant Model Context Protocol (MCP) tool definitions with correct schemas and built-in authentication.
The generated toolkits are deployed directly to the Kong AI Gateway, with policies enforced at runtime. Existing Kong security automatically extends to agent tooling.
Integration with broader AI infrastructure
Each toolkit generated by Context Mesh registers with the Kong Konnect MCP Registry for discovery and reuse across the enterprise. This gives developers a single place to find both APIs and agent tools. The system can define conditional policies and orchestration logic, including rate limiting, conditional routing, and data transformation.
By eliminating manual integration work, Context Mesh can help organizations reduce weeks of API inventory and mapping effort. It reuses existing Kong infrastructure to reduce costs and protects prior API investments during the transition to AI-first architectures.
Kong Context Mesh is now available in tech preview for Konnect deployments only.
Tip: Kong launches Konnect Metering & Billing for API monetization