gateway-workers

$npx mdskill add cloudflare/skills/gateway-workers

Read the current [Forward guide](https://developers.cloudflare.com/cloudflare-one/traffic-controls/integrate-workers/forward-traffic/index.md) before implementing a Worker. It defines how Gateway invokes Workers on permitted HTTP traffic. Retrieve the relevant page below rather than copying setup, API payloads, or limits into the project from memory.

SKILL.md

.github/skills/gateway-workersView on GitHub ↗
---
name: gateway-workers
description: Build, debug, or review Cloudflare Workers invoked by the Forward action in Gateway HTTP policies, including the GatewayMiddleware entrypoint, Gateway types, context, and continuation.
---

# Gateway Workers

Read the current [Forward guide](https://developers.cloudflare.com/cloudflare-one/traffic-controls/integrate-workers/forward-traffic/index.md) before implementing a Worker. It defines how Gateway invokes Workers on permitted HTTP traffic. Retrieve the relevant page below rather than copying setup, API payloads, or limits into the project from memory.

## Workflow

1. Inspect the existing Worker and choose a [Worker template](https://developers.cloudflare.com/cloudflare-one/traffic-controls/integrate-workers/worker-templates/index.md) for the required transformation, synthetic response, or background task.
2. Export a named `GatewayMiddleware` class extending `WorkerEntrypoint` from `cloudflare:workers`. Implement `handle(request, context, next)` and return a `Response`. Gateway calls this method through RPC. A default `fetch()` handler alone does not implement the Forward contract.
3. For TypeScript, follow the chosen template's current package installation and type-import instructions.
4. Use the second argument for [Gateway context](https://developers.cloudflare.com/cloudflare-one/traffic-controls/integrate-workers/available-context/index.md). It describes the original request and stays unchanged when the Worker modifies the request. Keep it distinct from `this.ctx` (Workers execution context) and `this.env` (bindings).
5. Continue the user's request with `next(request)` at most once, or return a synthetic response without calling it. `next()` always targets the original destination. Changes to the destination URL or host are ignored, and it cannot redirect, retry, or fan out. Use `fetch()` to send a request to a different destination. It creates an independent subrequest that does not inherit the user's Gateway policies or egress configuration.

## Context

Use the published types and context reference for field names and availability. Treat a falsy identity (`!identity`) as anonymous and guard optional posture values. Call `await context.identity()` or `await context.devicePosture()` only when needed. Both getters can reject. Catch lookup failures separately from missing values and decide how to handle each when the Worker depends on them. Use Gateway-provided identity rather than client-supplied identity headers.

## Gateway behavior

- Only the first matching Forward policy invokes a Worker. Worker invocation does not confirm successful body inspection. Follow the Forward guide's [inspection order](https://developers.cloudflare.com/cloudflare-one/traffic-controls/integrate-workers/forward-traffic/index.md#how-it-works).
- Gateway inspects the final response, including synthetic responses. Modified request bytes are not inspected again before reaching the origin.
- Review the Forward guide's [limitations](https://developers.cloudflare.com/cloudflare-one/traffic-controls/integrate-workers/forward-traffic/index.md#known-limitations) before choosing traffic: HTTPS requires decryption, WebSockets bypass the Worker, and matching Workers VPC traffic is dropped. Worker invocation failures fail closed.

## Validation

Type-check against the installed Gateway types and test the named RPC entrypoint, including zero or one continuation call and the context values the Worker uses. Then test a permitted request through a scoped Gateway Forward policy. A direct request to a default `fetch()` handler does not validate the Gateway path. Follow the chosen template's test procedure and the Forward guide's troubleshooting steps.

Check the [response and streaming rules](https://developers.cloudflare.com/cloudflare-one/traffic-controls/integrate-workers/forward-traffic/index.md#known-limitations): return a final status from `200` to `599`, return no body for HEAD requests or statuses `204`, `205`, and `304`, and stream bodies rather than buffering them.

## Related skills

- [cloudflare-one](../cloudflare-one/SKILL.md): Gateway policy design, traffic paths, identity, TLS, and rollout.
- [workers-best-practices](../workers-best-practices/SKILL.md): streaming, bindings, generated environment types, background work, observability, and runtime tests.
- [CLI selection](../cloudflare/SKILL.md#choose-the-cli): choose the existing `cf-cli` or `wrangler` skill for project setup and deployment.

## Skill feedback

If the user reports that this skill gave incorrect or outdated guidance, or is missing information, read [references/skill-feedback.md](references/skill-feedback.md).

More from cloudflare/skills

SkillDescription
agents-sdkBuild AI agents on Cloudflare Workers using the Agents SDK. Load when creating stateful agents, durable workflows, real-time WebSocket apps, scheduled tasks, MCP servers, or chat applications. Covers Agent class, state management, callable RPC, Workflows integration, and React hooks. Biases towards retrieval from Cloudflare docs over pre-trained knowledge.
agents-sdkBuild, debug, or review Cloudflare Agents SDK applications using the agents package.
basinBuild and troubleshoot Cloudflare Basin analytics workflows with Basin Pipelines, Basin Catalog, and Basin SQL. Use for streaming data into R2 Iceberg tables, managing catalogs, or querying those tables; also use for requests using the former Data Platform, Pipelines, R2 Data Catalog, or R2 SQL names.
building-ai-agent-on-cloudflare|
building-mcp-server-on-cloudflare|
cf-cliCloudflare's official cf CLI (beta) for the public Cloudflare API and Workers projects. Use for finding and running Cloudflare commands from the terminal — sign-in and profiles, cf cli search, cf schema, cf deploy/dev/build, resource commands for DNS, zones, D1, KV, and R2 — and for choosing between cf and Wrangler in a project.
cloudflareComprehensive Cloudflare platform skill covering Workers, Pages, storage (KV, D1, R2), AI (Workers AI, Vectorize, Agents SDK), networking (Tunnel, Spectrum), security (WAF, DDoS), and infrastructure-as-code (Terraform, Pulumi). Use for any Cloudflare development task. Biases towards retrieval from Cloudflare docs over pre-trained knowledge.
cloudflare-apiHit the Cloudflare REST API directly for operations that wrangler and MCP can't handle well. Bulk DNS, custom hostnames, email routing, cache purge, WAF rules, redirect rules, zone settings, Worker routes, D1 cross-database queries, R2 bulk operations, KV bulk read/write, Vectorize queries, Queues, and fleet-wide resource audits. Produces curl commands or scripts. Triggers: 'cloudflare api', 'bulk dns', 'custom hostname', 'email routing', 'cache purge', 'waf rule', 'd1 query', 'r2 bucket', 'kv bulk', 'vectorize query', 'audit resources', 'fleet operation'.
cloudflare-email-serviceImplement or troubleshoot Cloudflare Email Sending and Email Routing integrations and their delivery configuration.
cloudflare-oneDesign, configure, troubleshoot, or review Cloudflare One Zero Trust and SASE deployments. Use cloudflare-one-migrations for migration planning from other vendors.