https://praxis.fast/grid/v0.2.2/Recent content in AI Grid Network (AGN) Documentation on PraxisHugoenhttps://praxis.fast/grid/v0.2.2/installation/migration-0.3.x/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/installation/migration-0.3.x/The 0.3.x release accepts both the previous and current names for CRD spec fields.https://praxis.fast/grid/v0.2.2/adding-provider/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/adding-provider/Step-by-step workflow for adding a new inference provider to an existing AGN installation. Covers in-cluster backends, existing services, and external HTTPS APIs.https://praxis.fast/grid/v0.2.2/architecture/provider-draining/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/architecture/provider-draining/AGN supports graceful maintenance of inference providers through the optional InferenceProvider.spec.trafficPolicy.drain field.https://praxis.fast/grid/v0.2.2/architecture/operations/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/architecture/operations/This guide covers AGN production installation, network formation, site lifecycle, routing configuration, security, and observability.https://praxis.fast/grid/v0.2.2/architecture/routing/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/architecture/routing/Describes how Grid renders, versions, distributes, and secures routing state for Praxis gateways.https://praxis.fast/grid/v0.2.2/routing/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/routing/AI Grid Network (AGN) routes inference traffic across provider gateways using a multi-dimensional policy rather than a single load-balancing algorithm.https://praxis.fast/grid/v0.2.2/architecture/overview/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/architecture/overview/AI Grid Network (AGN) is the control plane that prepares routing state for Praxis AI gateways.https://praxis.fast/grid/v0.2.2/architecture/auth/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/architecture/auth/AI Grid Network (AGN) deployments involve up to three distinct authentication layers. Each serves a different trust boundary and must not be conflated.https://praxis.fast/grid/v0.2.2/architecture/ci-kind-e2e/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/architecture/ci-kind-e2e/This document describes the CI coverage for AGN’s Kind-based end-to-end validation suite.https://praxis.fast/grid/v0.2.2/site-selection/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/site-selection/How a grid gateway serving gridServing chooses a site for each request in grid_site_route, what feeds the choice, and what you can change.https://praxis.fast/grid/v0.2.2/architecture/crds/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/architecture/crds/API group: grid.praxis.fast/v1alpha1https://praxis.fast/grid/v0.2.2/development/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/development/Development and validation guidance for AI Grid Network (AGN).https://praxis.fast/grid/v0.2.2/developing/conventions/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/developing/conventions/Points to the canonical Grid development conventions for code, tests, documentation, and pull requests.https://praxis.fast/grid/v0.2.2/conventions/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/conventions/This is the canonical development policy for AI Grid Network (AGN). Thehttps://praxis.fast/grid/v0.2.2/architecture/enrollment/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/architecture/enrollment/Enrollment is how a cluster, from then on a site, gets its grid identity: a name the grid assigned and a certificate the grid CA signed.https://praxis.fast/grid/v0.2.2/installation/existing-clusters/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/installation/existing-clusters/This Helm-based installation workflow installs the AGN Operator and Praxis gateways on existing Kubernetes clusters.https://praxis.fast/grid/v0.2.2/architecture/external-ingress/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/architecture/external-ingress/External client ingress extends AGN’s workload-routing model to a stable public endpoint. A global traffic manager selects an edge. AGN and Praxis then select and reach an eligible provider.https://praxis.fast/grid/v0.2.2/fips/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/fips/For a FIPS build, the security-relevant crypto (the TLS stacks and certificate identity hashing) must route through the system openssl (dynamically linked libssl and libcrypto).https://praxis.fast/grid/v0.2.2/developing/forge/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/developing/forge/Install and use the pinned praxis-forge CLI for Grid topology validation and qualification workflows.https://praxis.fast/grid/v0.2.2/installation/readme/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/installation/readme/Introduces Grid installation paths for existing Kubernetes clusters, site enrollment, and AWS deployments.https://praxis.fast/grid/v0.2.2/installation/aws/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/installation/aws/Plan multi-site Grid deployments on AWS, including private connectivity, load balancers, and chart settings.https://praxis.fast/grid/v0.2.2/metrics/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/metrics/Every metric the grid exports, what it means, and where to scrape it. Three components expose Prometheus endpoints and they answer different questions.https://praxis.fast/grid/v0.2.2/architecture/opentelemetry/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/architecture/opentelemetry/Grid gateway builds can export traces to an OTLP/gRPC collector. Export is configured on the gateway process; routing overlays remain limited to routing state.https://praxis.fast/grid/v0.2.2/architecture/consumer-config/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/architecture/consumer-config/The AGN Operator can generate the consumer Praxis ConfigMap from routing overlay data. This is an opt-in feature on each GatewayRef.https://praxis.fast/grid/v0.2.2/architecture/polling-metrics/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/architecture/polling-metrics/A gateway prefers the least-loaded site serving a model, so it needs each site’s load.https://praxis.fast/grid/v0.2.2/architecture/prefix-affinity/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/architecture/prefix-affinity/The grid gateway’s grid_site_route filter keeps a conversation on the site that already holds its prompt in the prefix cache, while that site’s queue allows.https://praxis.fast/grid/v0.2.2/architecture/scoring/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/architecture/scoring/For how scoring interacts with routing policy, selection groups, and request-time selection, see the AGN Routing Guide.https://praxis.fast/grid/v0.2.2/release/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/release/AI Grid Network (AGN) uses Semantic Versioning.https://praxis.fast/grid/v0.2.2/developing/unicode-audit/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/developing/unicode-audit/The audit covers the tracked repository text after integrating main at af82146cf9d23fa1fa5d7ffb587fb5053ca23a92.https://praxis.fast/grid/v0.2.2/architecture/signals/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/architecture/signals/Signals are the provider load a site observes, such as queue depth and KV-cache pressure.https://praxis.fast/grid/v0.2.2/installation/enrollment/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/installation/enrollment/A site enrolls with a grid by redeeming a one-time token at the enrollment service.https://praxis.fast/grid/v0.2.2/technical-preview-scope/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/technical-preview-scope/AI Grid Network (AGN) is a network-routing component for AI traffic.https://praxis.fast/grid/v0.2.2/developing/type-design/Mon, 01 Jan 0001 00:00:00 +0000https://praxis.fast/grid/v0.2.2/developing/type-design/Make invalid states unrepresentable. The type system and serde should enforce constraints at parse time, not at runtime.