Ross ROSS = Recommend OSS · open-source software intelligence for agents

agentgateway/agentgateway

Next Generation Agentic Proxy for AI Agents and MCP servers observed · 2026-08-28

github.com/agentgateway/agentgateway · homepage · Rust · Apache-2.0 (permissive) observed · 2026-08-28

Health v2 · maintenance only

81/100

  • Activity 99
  • Release rhythm 83
  • Longevity 38
How is this computed?

round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10) — computed 2026-09-03. Adoption (stars, forks) is never an input.

  • gap_med: 2.0
  • age_days: 533
  • days_rel: 35
  • days_push: 7
  • n_releases_24m: 91

Full methodology

Adoption not part of the score

4563 stars · 770 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

Agentgateway is an open-source, Rust-based proxy and gateway for agentic AI traffic, handling LLM provider routing, MCP tool servers, and A2A agent-to-agent communication alongside traditional HTTP, gRPC, and TCP traffic in one data plane. It can run as a standalone binary or as a Kubernetes Gateway API data plane with a built-in control plane, offering security, observability, and cost governance.

Use cases

  • proxy LLM requests to OpenAI, Anthropic, Gemini, and Bedrock through one OpenAI-compatible API
  • expose and federate MCP tool servers to AI clients with OAuth and authorization
  • secure and observe agent-to-agent (A2A) communication
  • enforce LLM spend budgets, rate limits, and prompt guardrails
  • load balance and fail over across multiple LLM providers
  • route inference requests to self-hosted model pools on Kubernetes
  • front regular APIs and microservices with the same gateway used for AI traffic

When to choose

  • you need a single gateway for both traditional service traffic and AI-native protocols (MCP, A2A, LLM providers)
  • you run Kubernetes and want Gateway API / Inference Extension integration
  • you need enterprise features like mTLS, OIDC, multi-tenancy, and cost attribution for agent workloads
  • you want tool federation and policy enforcement in front of MCP servers

When to avoid

  • you only need a simple reverse proxy with no AI-specific routing or policy needs
  • you want a lightweight client-side SDK rather than an infrastructure gateway
  • your stack has no Kubernetes or container deployment capability and you need a managed SaaS gateway

Facets

service · maturity active

api-gateway proxy llm-inference mcp agent-framework rate-limiting auth authorization monitoring middleware routing websocket rpc artificial-intelligence large-language-models apis microservices cloud-computing security networking self-hosted rust self-hosted cloud cli cross-platform mcp-gateway a2a llm-gateway gateway-api inference-routing service-mesh openai-compatible ai-gateway traffic-management cost-controls ai-agents kubernetes docker

10 sources

Member repositories

RepositoryRoleHealth v2
agentgateway/agentgatewaymain81

For agents

markdown · JSON · MCP: product_card(name="agentgateway/agentgateway")

Data as of 2026-08-30T08:39:29.467469+00:00 · Report a problem