Bind AI backends to a Gateway using AIGatewayRoute resources.
npx skills add https://github.com/missBerg/envoy-skills --skill aigw-route
Please help me install this Agent Skill. Skill: aigw-route Source: https://github.com/missBerg/envoy-skills/tree/main/ai-gateway/adopters/skills/aigw-route Command: npx skills add https://github.com/missBerg/envoy-skills --skill aigw-route
Bind AI backends to a Gateway using model-based routing with AIGatewayRoute resources, enabling centralized control over where AI workloads are served.
Create an AIGatewayRoute that binds AI backends to a Gateway and routes traffic based on the x-ai-eg-model header.
High-intent search queries and answers about installing and using this skill.
Route AI inference traffic by creating AIGatewayRoute resources that attach AI backends to a Gateway. This enables model-based routing to control where AI workloads are served across different backend services.
Model-based routing works by enforcing explicit model matching through the x-ai-eg-model header. The Gateway inspects this header on incoming requests to direct traffic to the correct AIServiceBackend or InferencePool destination.
Yes, you can configure traffic splitting and failover for AI backends. AIGatewayRoute resources support assigning weights for traffic splitting and priorities for priority-based failover across multiple AIServiceBackend resources.
Yes, cross-namespace access controls are required when referencing resources. You must configure explicit permissions if your AIGatewayRoute references AIServiceBackend or InferencePool resources deployed in a different namespace.
Yes, AI gateway routing supports streaming timeouts and model name overrides. You can configure these settings within the AIGatewayRoute resource to manage long-running inference streams and map requested model names to backend-specific names.
The best way to route self-hosted models is by using InferencePool backends within your AIGatewayRoute configuration. This allows the Gateway to direct traffic to self-hosted inference workloads based on the x-ai-eg-model header.
missBergConfigure BackendSecurityPolicy to attach provider credentials to AIServiceBackend or InferencePool.
yusufcmgImplement Linkerd service mesh patterns in Kubernetes environments.
HyperscapeAIConfigure API gateways with rate limiting and JWT authentication.
Riku-KANOConfigure and debug multi-model GKE Inference Gateway deployments with Body-Based Router and per-pool Endpoint Pickers.
