# Fabric Gateway Docs | Documentation Instructions for AI Agents: - For clean Markdown of any page, append `.md` to the page URL - For section-specific indexes, append `/llms.txt` to any section URL - For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.fabricgateway.ai/_mcp/server ## Docs - [Welcome to Fabric AI Gateway docs](https://docs.fabricgateway.ai/get-started/overview.md): Learn how to connect, secure, route, and monitor LLM, MCP, and API traffic with Fabric AI Gateway. - [Explore the console](https://docs.fabricgateway.ai/get-started/console.md): Find out which Fabric AI Gateway console items you can see, and set up your first route. - [Quickstart: Send your first request on the hosted gateway](https://docs.fabricgateway.ai/get-started/quickstart.md): Connect your first LLM provider or MCP server, and send a test request through the gateway. - [Quickstart: Run a data plane in your VPC](https://docs.fabricgateway.ai/get-started/quickstart-data-plane.md): Install the Fabric AI Gateway data plane in your Kubernetes cluster, connect it to the control plane with workload attestation, and check that it serves traffic. - [What is Fabric AI Gateway](https://docs.fabricgateway.ai/evaluate/what-is-fabric.md): Decide whether Fabric AI Gateway fits your team. See what it covers, how it runs, what it protects, and how to try it. - [Understand who uses Fabric](https://docs.fabricgateway.ai/evaluate/who-uses-fabric.md): Find the role that matches your job, and go straight to the part of the docs that covers what you do. - [How Fabric works](https://docs.fabricgateway.ai/evaluate/architecture.md): See how the control plane and data plane divide the work, what each component does, and how a request moves through the gateway. - [Choose a deployment model](https://docs.fabricgateway.ai/evaluate/deployment-models.md): Compare the Free (hosted) and Enterprise (hybrid) plans, what runs where, what you operate, and how the data plane stays connected to the control plane. - [Review security and data handling](https://docs.fabricgateway.ai/evaluate/security-and-data-handling.md): See where traffic and secrets go, what Postman can access, and how the gateway behaves when something fails. - [Know the current limitations](https://docs.fabricgateway.ai/evaluate/limits.md): Check the boundaries of Fabric AI Gateway before you commit, including deployment options, API translation, and access approval. - [Try Fabric](https://docs.fabricgateway.ai/evaluate/try-fabric.md): Pick the fastest way to test Fabric AI Gateway with your own providers, and find out how to talk to Postman about a larger evaluation. - [Learn the key concepts](https://docs.fabricgateway.ai/concepts/key-concepts.md): See how the main parts of Fabric AI Gateway fit together, and find the definition of any term in the Glossary. - [Understand credentials and keys](https://docs.fabricgateway.ai/concepts/credentials-and-keys.md): Tell apart the four ways that people and systems authenticate with Fabric AI Gateway, and find the topic for each. - [Follow a request through the gateway](https://docs.fabricgateway.ai/concepts/life-of-a-request.md): Follow a request through the data plane, from route matching and authentication through the plugin chain and the upstream call to logging and metering. - [Use the gateway](https://docs.fabricgateway.ai/use.md): Send requests through the gateway from your apps, agents, and tools. - [Authenticate and send a request](https://docs.fabricgateway.ai/use/authenticate.md): Find the gateway's base URL, send your gateway key in the fg-api-key header, and make your first request through Fabric AI Gateway. - [Discover what's available](https://docs.fabricgateway.ai/use/discover.md): Find out which models, MCP servers, and APIs your gateway exposes, which path reaches each one, and whether it needs a gateway key. - [Choose a model](https://docs.fabricgateway.ai/use/llm/choose-a-model.md): Name a model in your request, list the models a route serves, and understand the errors for a model the gateway can't serve. - [Use the OpenAI-compatible API](https://docs.fabricgateway.ai/use/llm/openai-compatible.md): Call chat completions, responses, embeddings, and the model list through an LLM route, with streaming and OpenAI-shaped errors. - [Use the Anthropic Messages API](https://docs.fabricgateway.ai/use/llm/anthropic-messages.md): Call messages, token counting, and the model list through an LLM route, with streaming and Anthropic-shaped errors. - [Connect an MCP client to an MCP route](https://docs.fabricgateway.ai/use/mcp/connect-a-client.md): Point an MCP client at an MCP route, authenticate with a gateway key or by signing in, and check that the client sees the route's tools. - [Connect your accounts to MCP servers](https://docs.fabricgateway.ai/use/mcp/connect-your-accounts.md): Connect your own account on an upstream service, such as GitHub or Atlassian, so the gateway can call that service's MCP server for you. - [Call a REST API through the gateway](https://docs.fabricgateway.ai/use/call-a-rest-api.md): Send requests to a REST API route, authenticate with your gateway key, and learn what the gateway changes on the way to the API. - [Troubleshoot access](https://docs.fabricgateway.ai/use/troubleshoot-access.md): Work out which step refused your request — the gateway key, access control, or the upstream credential — and who can fix it. - [Handle errors and retries](https://docs.fabricgateway.ai/use/errors-and-retries.md): Recognize gateway errors on REST, LLM, and MCP routes, decide which ones to retry, and read rate-limit responses and timeouts. - [See your requests](https://docs.fabricgateway.ai/use/see-your-requests.md): Find a request in traces, check which model or provider answered, and give your admin the details they need. - [Monitor gateway traffic](https://docs.fabricgateway.ai/use/monitor.md): Review aggregated analytics and inspect individual requests that pass through Fabric AI Gateway. - [Review analytics](https://docs.fabricgateway.ai/use/monitor/analytics.md): Review traffic, errors, latency, token usage, and cost for LLM, MCP, and HTTP traffic over a time window. - [Inspect traces](https://docs.fabricgateway.ai/use/monitor/traces.md): Find and inspect distributed traces for LLM, MCP, and HTTP traffic. - [Sign in to fabric-observability](https://docs.fabricgateway.ai/use/monitor/sign-in-to-fabric-observability.md): Generate a one-time token in the console to sign in to your organization's fabric-observability app, and fix the errors that can stop it. - [Configure the gateway](https://docs.fabricgateway.ai/configure.md): Find the right topic for connecting providers, routing and securing traffic, monitoring requests, and managing access in Fabric AI Gateway. - [Manage connections and routes](https://docs.fabricgateway.ai/configure/connections.md): Connect and manage LLM providers, MCP servers, and REST APIs, and review the routes that expose them. - [Connect an LLM provider](https://docs.fabricgateway.ai/configure/connections/llm-providers.md): Connect an LLM provider, create a route for it, and create a gateway key for callers. - [Connect to a provider over a private network](https://docs.fabricgateway.ai/configure/connections/private-connections.md): Send LLM traffic to a provider through Fabric's private endpoint instead of the public internet, and set up the AWS role that Fabric assumes. - [Connect an MCP server](https://docs.fabricgateway.ai/configure/connections/mcp-servers.md): Create an MCP server connection, add upstreams, and set the access mode. - [Add an OAuth client for an MCP server](https://docs.fabricgateway.ai/configure/connections/oauth-clients.md): Register your own OAuth app with Fabric for an MCP server that doesn't support dynamic client registration, and use it on an MCP connection. - [Connect a REST API](https://docs.fabricgateway.ai/configure/connections/rest-apis.md): Connect a REST API as a connection, either manually or by importing an OpenAPI spec that generates its routes. - [Connect OpenAI](https://docs.fabricgateway.ai/configure/providers/openai.md): Target URL, credentials, supported endpoints, and model names for an OpenAI connection. - [Connect Anthropic](https://docs.fabricgateway.ai/configure/providers/anthropic.md): Target URL, credentials, supported endpoints, and model names for an Anthropic connection. - [Connect Amazon Bedrock](https://docs.fabricgateway.ai/configure/providers/amazon-bedrock.md): Target URLs, AWS credentials, supported endpoints, and model IDs for Amazon Bedrock and Amazon Bedrock Mantle connections. - [Connect Azure OpenAI](https://docs.fabricgateway.ai/configure/providers/azure-openai.md): Target URL, credentials, supported endpoints, and deployment names for an Azure OpenAI connection. - [Connect Google Gemini](https://docs.fabricgateway.ai/configure/providers/google-gemini.md): Target URL, credentials, supported endpoints, and model names for a Google Gemini AI Studio connection. - [Connect an OpenAI-compatible or self-hosted model server](https://docs.fabricgateway.ai/configure/providers/openai-compatible.md): Connect any server that implements the OpenAI or Anthropic Messages API, including models you host yourself. - [Other LLM providers](https://docs.fabricgateway.ai/configure/providers/other-providers.md): How to connect LLM providers that don't have their own page, including Google Vertex AI. - [Manage credentials and vaults](https://docs.fabricgateway.ai/configure/credentials.md): Manage the credentials that the gateway uses, and understand where secrets are stored on cloud and enterprise tenants. - [Route LLM requests across models and providers](https://docs.fabricgateway.ai/configure/routing.md): Compare the routing plugins that choose a model for each LLM request or fail over across providers. - [Choose a routing approach](https://docs.fabricgateway.ai/configure/routing/choose-an-approach.md): Compare the model router, the LLM Router, and the Adaptive Model Router, and pick the one that fits a route. - [Route requests by model name](https://docs.fabricgateway.ai/configure/routing/model-router.md): Serve many providers and models from one route, and let clients pick one by naming it as provider/model. - [Define model aliases](https://docs.fabricgateway.ai/configure/routing/aliases.md): Give clients a plain model name that maps to one or more providers, and retarget it without changing any client. - [Route requests with the Adaptive Model Router](https://docs.fabricgateway.ai/configure/routing/adaptive-routing.md): Rate each LLM request with a decision service and send it to the model that fits its difficulty. - [Fail over across providers with the LLM Router](https://docs.fabricgateway.ai/configure/routing/failover.md): List connections in priority order so that a request that fails on one provider moves to the next. - [Route requests through Lightyear](https://docs.fabricgateway.ai/configure/routing/lightyear.md): Use the Lightyear mode of the LLM Router plugin to route requests through Lightyear and reduce LLM spend. - [Authenticate requests with Key Auth](https://docs.fabricgateway.ai/configure/authentication/key-auth.md): Require a gateway key on a route, choose the header that carries it, and troubleshoot authentication errors. - [Authenticate requests with JWT Auth](https://docs.fabricgateway.ai/configure/authentication/jwt-auth.md): Authenticate callers with JSON Web Tokens from your identity provider, and use token claims to control which tools an agent can use. - [Restrict model and tool access](https://docs.fabricgateway.ai/configure/access-and-limits/model-access.md): Limit which models a connection serves, and which users can use a connection and its tools. - [Limit requests and tokens](https://docs.fabricgateway.ai/configure/access-and-limits/rate-limits.md): Cap how many requests or LLM tokens a caller can use in a time window, and what clients get when they reach the limit. - [Control spend with budgets](https://docs.fabricgateway.ai/configure/access-and-limits/budgets.md): Set a spending budget for your organization or a namespace, adjust costs for your provider discounts, cap what you pay Fabric, and cap token use per caller. - [Control traffic with plugins](https://docs.fabricgateway.ai/configure/plugins.md): Secure, transform, observe, and control the cost of LLM, MCP, and HTTP API traffic with plugins. - [Administer your organization](https://docs.fabricgateway.ai/configure/organization.md): Understand the roles that control access to your Fabric AI Gateway organization and its namespaces, how cloud and enterprise tenants differ, and where to review activity and billing. - [Manage users and invitations](https://docs.fabricgateway.ai/configure/organization/users.md): Review users, invite people, check pending invitations, and change a user's role. - [Manage namespaces](https://docs.fabricgateway.ai/configure/organization/namespaces.md): Give each team its own space for routes, connections, credentials, analytics, traces, and people within one organization, and track what each team spends. - [Manage API keys](https://docs.fabricgateway.ai/configure/automation/api-keys.md): Create admin API keys to manage your Fabric AI Gateway organization programmatically. - [Run the data plane in your VPC](https://docs.fabricgateway.ai/run.md): What the Fabric AI Gateway data plane is, what you run versus what Postman runs, what the Helm chart installs, and how the data plane connects to Postman. - [Requirements and sizing](https://docs.fabricgateway.ai/run/plan-a-deployment.md): Check the Kubernetes, capacity, and external dependencies the Fabric AI Gateway data plane needs before you install it in your VPC. - [Network and egress](https://docs.fabricgateway.ai/run/network.md): Every network path the Fabric AI Gateway data plane uses, the egress you need to allow, and the ports inside the cluster. - [Install the data plane with Helm](https://docs.fabricgateway.ai/run/install.md): Install the Fabric AI Gateway data plane in your Kubernetes cluster with Helm, see what Helm does in order, and verify the install. - [Database migrations](https://docs.fabricgateway.ai/run/database-migrations.md): How the data plane chart migrates its MySQL schema, and how to adopt an existing database, run migrations yourself, or use a separate migration user. - [Connect the data plane to the control plane](https://docs.fabricgateway.ai/run/connect-a-data-plane.md): Enroll the data plane with Kubernetes workload attestation or a bootstrap token, supply the control plane CA certificate, and check the data plane in the console. - [Expose the gateway to your clients](https://docs.fabricgateway.ai/run/expose-the-gateway.md): Make the Fabric AI Gateway data plane reachable from outside the cluster with an ingress or load balancer, end TLS, and set its host name in your values and the console. - [Configure a secret store](https://docs.fabricgateway.ai/run/configure-a-secret-store.md): Choose the secret store and vault for a data plane you run, set up Azure Key Vault with managed identity or HashiCorp Vault, and fix common errors. - [Data plane production checklist](https://docs.fabricgateway.ai/run/production-checklist.md): Check your data plane's security, availability, and observability settings against the chart's production defaults before it carries production traffic. - [Helm values](https://docs.fabricgateway.ai/run/helm-values.md): Reference for the data plane Helm chart values, environment variable layering, feature toggles, and the defaults to retarget before you install. - [Secrets and keys](https://docs.fabricgateway.ai/run/secrets-and-keys.md): Create the Kubernetes Secret the data plane reads, generate its encryption key, supply the control plane CA certificate, and rotate each value. - [Monitor the data plane](https://docs.fabricgateway.ai/run/monitor.md): Check data plane health with kubectl, its health endpoints, and the console, and collect its logs, metrics, and traces in your own observability stack. - [Scale and run a highly available data plane](https://docs.fabricgateway.ai/run/scale-and-high-availability.md): Run several gateway replicas, add autoscaling, keep node drains working, choose a secret store, and understand what the data plane does when the control plane is unreachable. - [Run the tool-search embedder](https://docs.fabricgateway.ai/run/embedder.md): Understand the embedder the data plane chart deploys for MCP tool search, and keep its model cache, scaling, drains, and network access under control. - [Troubleshoot the data plane](https://docs.fabricgateway.ai/run/troubleshooting.md): Find the cause and fix for common data plane problems, from failed migrations and crashing pods to enrollment errors, missing telemetry, and embedder issues. - [Maintain the data plane and its data](https://docs.fabricgateway.ai/run/maintenance.md): Learn what state the data plane keeps and where, what you need to back up, how long stored data lasts, and the routine tasks that keep a data plane healthy. - [Upgrade the data plane](https://docs.fabricgateway.ai/run/upgrade.md): Upgrade the data plane's chart and image with Helm, understand what runs during an upgrade, and roll back if you need to. - [Data plane chart release notes](https://docs.fabricgateway.ai/run/release-notes.md): See what changed in each version of the data plane Helm chart, and what to do when you upgrade to it. - [Uninstall the data plane](https://docs.fabricgateway.ai/run/uninstall.md): Remove the data plane Helm release, and clean up the Secrets, volumes, database, and console nodes that the uninstall leaves behind. - [Troubleshoot Fabric AI Gateway](https://docs.fabricgateway.ai/help/troubleshooting.md): Find the cause of common errors when you send requests, configure routes, or run a data plane, and how to fix them. - [Frequently asked questions](https://docs.fabricgateway.ai/help/faq.md): Answers to common questions about sending requests through Fabric AI Gateway, routing, errors, and deployment. - [Glossary](https://docs.fabricgateway.ai/help/glossary.md): Definitions of the terms used in the Fabric AI Gateway console and docs. - [Use Claude Code with Fabric AI Gateway](https://docs.fabricgateway.ai/guides/coding-agents/claude-code.md): Send Claude Code traffic through a Fabric AI Gateway route and confirm that requests arrive. - [Use Codex with Fabric AI Gateway](https://docs.fabricgateway.ai/guides/coding-agents/codex.md): Send Codex traffic through a Fabric AI Gateway route and confirm that requests arrive. - [Use OpenCode with Fabric AI Gateway](https://docs.fabricgateway.ai/guides/coding-agents/opencode.md): Send OpenCode traffic through a Fabric AI Gateway route and confirm that requests arrive. - [Use the OpenAI SDK with Fabric AI Gateway](https://docs.fabricgateway.ai/guides/sdks/openai.md): Point the OpenAI SDK at a Fabric AI Gateway route and send your gateway key in the fg-api-key header. - [Use the Anthropic SDK with Fabric AI Gateway](https://docs.fabricgateway.ai/guides/sdks/anthropic.md): Point the Anthropic SDK at a Fabric AI Gateway route and send your gateway key in the fg-api-key header. - [Use the Vercel AI SDK with Fabric AI Gateway](https://docs.fabricgateway.ai/guides/sdks/vercel-ai.md): Point the Vercel AI SDK's OpenAI and Anthropic providers at a Fabric AI Gateway route and send your gateway key in the fg-api-key header. - [Export telemetry with OpenTelemetry](https://docs.fabricgateway.ai/guides/observability/opentelemetry.md): Send traces, metrics, and logs from an Enterprise data plane to your own OpenTelemetry backend, tune what it exports, or turn export off. - [Send data plane telemetry to Datadog](https://docs.fabricgateway.ai/guides/observability/datadog.md): Send traces, metrics, and logs from an Enterprise data plane to Datadog through a Datadog Agent's OTLP intake. - [Send data plane telemetry to Grafana Cloud](https://docs.fabricgateway.ai/guides/observability/grafana-cloud.md): Send traces, metrics, and logs from an Enterprise data plane to Grafana Cloud through a collector sidecar that re-exports over OTLP/HTTP. - [Send data plane telemetry to Honeycomb](https://docs.fabricgateway.ai/guides/observability/honeycomb.md): Send traces, metrics, and logs from an Enterprise data plane straight to Honeycomb over OTLP/gRPC. - [Send data plane telemetry to New Relic](https://docs.fabricgateway.ai/guides/observability/new-relic.md): Send traces, metrics, and logs from an Enterprise data plane straight to New Relic over OTLP/gRPC. - [Gateway API reference](https://docs.fabricgateway.ai/reference/gateway-api.md): The endpoints that your applications call on the Fabric AI Gateway data plane, for LLM, MCP, and REST routes, and the endpoints the gateway answers itself. - [Supported endpoints](https://docs.fabricgateway.ai/reference/supported-endpoints.md): Which LLM endpoints each kind of provider connection supports in Fabric AI Gateway, and what the gateway does with each endpoint. - [Request and response headers](https://docs.fabricgateway.ai/reference/headers.md): The request headers that Fabric AI Gateway reads, the response headers it adds, and the headers it changes on the way to the upstream. - [Gateway error codes](https://docs.fabricgateway.ai/reference/error-codes.md): Every error that Fabric AI Gateway returns itself, with its HTTP status, body, cause, and what to do. - [Limits and defaults](https://docs.fabricgateway.ai/reference/limits.md): Body size caps, timeouts, retry and health-check defaults, and streaming limits in Fabric AI Gateway. - [Plugin catalog](https://docs.fabricgateway.ai/reference/plugins.md): Browse the plugins that you can attach to routes, connections, and consumers in Fabric AI Gateway. - [Key Auth plugin](https://docs.fabricgateway.ai/reference/plugins/key-auth.md): Reference for the Key Auth plugin, which requires a gateway key in the fg-api-key header. - [JWT Auth plugin](https://docs.fabricgateway.ai/reference/plugins/jwt-auth.md): Reference for the JWT Auth plugin, which validates JSON Web Tokens from your identity provider. - [Access Control plugin](https://docs.fabricgateway.ai/reference/plugins/access-control.md): Reference for the Access Control plugin, which grants a consumer access to connections, methods, paths, and MCP tools. - [Deny All plugin](https://docs.fabricgateway.ai/reference/plugins/deny-all.md): Reference for the Deny All plugin, which refuses every request that it applies to. - [Validator plugin](https://docs.fabricgateway.ai/reference/plugins/validator.md): Reference for the Validator plugin, which checks request and response fields against rules. - [PII Redaction plugin](https://docs.fabricgateway.ai/reference/plugins/pii-redaction.md): Reference for the PII Redaction plugin, which masks, hashes, or blocks personal data in LLM prompts and MCP messages. - [Agent Scope Gate plugin](https://docs.fabricgateway.ai/reference/plugins/agent-scope-gate.md): Reference and setup steps for the Agent Scope Gate plugin, which requires agents to hold scopes, a principal type, or a consumer group. - [Agent Static Deny plugin](https://docs.fabricgateway.ai/reference/plugins/agent-static-deny.md): Reference and setup steps for the Agent Static Deny plugin, which blocks requests that match fixed rules. - [Agent Approval plugin](https://docs.fabricgateway.ai/reference/plugins/agent-approval.md): Reference and setup steps for the Agent Approval plugin, which requires a human approval token before an agent can make matching requests. - [Agent Audit plugin](https://docs.fabricgateway.ai/reference/plugins/agent-audit.md): Reference and setup steps for the Agent Audit plugin, which writes an audit event for each request from a caller with an agent identity. - [Passthrough plugin](https://docs.fabricgateway.ai/reference/plugins/passthrough.md): Reference for the Passthrough plugin, which lets callers reach an LLM provider with their own credentials through a route with no connection. - [LLM Observability plugin](https://docs.fabricgateway.ai/reference/plugins/llm-observability.md): Reference for the LLM Observability plugin, which records token usage, model, and finish reason for every LLM request. - [Tool Search plugin](https://docs.fabricgateway.ai/reference/plugins/tool-search.md): Reference for the Tool Search plugin, which replaces a large MCP tool list with three tools that search, describe, and call the real tools. - [MCP HTTP Bridge plugin](https://docs.fabricgateway.ai/reference/plugins/mcp-http-bridge.md): Reference and setup steps for the MCP HTTP Bridge plugin, which serves a REST API on an MCP route as MCP tools generated from its OpenAPI spec. - [LLM Budget plugin](https://docs.fabricgateway.ai/reference/plugins/llm-budget.md): Reference and setup steps for the LLM Budget plugin, which caps LLM spending in US dollars for a week or a month. - [Rate Limit plugin](https://docs.fabricgateway.ai/reference/plugins/rate-limit.md): Reference for the Rate Limit plugin, which limits the number of requests in a time window. - [LLM Rate Limit plugin](https://docs.fabricgateway.ai/reference/plugins/llm-rate-limit.md): Reference for the LLM Rate Limit plugin, which limits the LLM tokens a caller can use in a time window. - [Transformer plugin](https://docs.fabricgateway.ai/reference/plugins/transformer.md): Reference for the Transformer plugin, which changes headers, query parameters, cookies, and JSON body fields. - [Secret Injector plugin](https://docs.fabricgateway.ai/reference/plugins/secret-injector.md): Reference for the Secret Injector plugin, which adds a secret from a vault to upstream requests. - [LLM Router plugin](https://docs.fabricgateway.ai/reference/plugins/llm-router.md): Reference for the LLM Router plugin, which sends a route's LLM requests to one connection, a fallback chain, or a weighted mix. - [LLM Model Router plugin](https://docs.fabricgateway.ai/reference/plugins/llm-model-router.md): Reference for the model router (`llm-model-router`) plugin, which routes a request to the connections that serve the model it names. - [Adaptive Model Router plugin](https://docs.fabricgateway.ai/reference/plugins/adaptive-model-router.md): Reference for the Adaptive Model Router plugin, which picks a model and reasoning effort for each LLM request. - [Changelog](https://docs.fabricgateway.ai/changelog/llms.txt)