How do you customize request headers or authentication tokens in RTK Query?
In RTK Query, you can easily customize HTTP request headers — including adding authentication tokens — using the fetchBaseQuery options or a custom base query. This allows you to attach tokens dynamically from your Redux store or local storage before each request.
- 1
Using
prepareHeaders: Modify headers before each request by accessing the Redux state or adding authentication tokens. - 2
Using a Custom Base Query: Create your own base query wrapper to handle special authentication logic, token refresh, or custom header rules.
- 3
Endpoint-Level Overrides: Add or modify headers directly within individual endpoints when specific routes require custom headers.
In this example, the prepareHeaders function adds an Authorization header for each request. It retrieves the token from the Redux store and ensures that every endpoint automatically uses it.
This approach gives full control over how requests and responses are handled, making it ideal for implementing token refresh workflows or handling special authentication rules.
By customizing headers with prepareHeaders or a custom base query, RTK Query enables secure and flexible communication with authenticated APIs.
Scenario Questions
0-2 years experience
- 1We have a simple RTK Query API slice that fetches a list of products. How would you add an Authorization header with a static token to every request?
- 2If the backend expects a custom header X-Client-Version, where in the RTK Query setup would you inject that header, and what code would you write?
- 3What happens if you forget to return the modified headers object in the prepareHeaders callback?
2-5 years experience
- 1Your app needs to include a JWT that can change after login. Describe how you would configure RTK Query to attach the latest token to each request, and how you would handle token refresh failures.
- 2During a recent feature rollout, some API calls started failing with 401 errors after you added a new header. Walk me through how you would debug whether the issue is in the header injection or the server.
- 3You need to send different authentication schemes (Bearer token for some endpoints, API key for others) within the same RTK Query service. How would you structure the baseQuery or endpoints to support this?
5-8 years experience
- 1At scale, our service makes thousands of concurrent RTK Query calls, each needing per‑request correlation IDs. How would you design a solution to add a unique X-Request-ID header efficiently without leaking memory or causing race conditions?
- 2Explain the trade‑offs between using a custom baseQuery that injects headers versus using endpoint‑level prepareHeaders calls, especially regarding code maintainability and testability.
- 3If we need to rotate authentication tokens automatically and ensure in‑flight requests use the old token while new requests use the refreshed one, how would you implement this in RTK Query without causing request duplication or lost responses?
8+ years experience
- 1Our organization is moving from a monolithic auth service to a distributed zero‑trust model, requiring per‑service signed JWTs. How would you evolve the existing RTK Query header handling across multiple micro‑frontends to support this change while keeping backward compatibility?
- 2Discuss how you would standardize header and token management across dozens of RTK Query services in a large codebase, including tooling, lint rules, and runtime validation, to prevent security regressions.
- 3When integrating RTK Query with a legacy API gateway that expects signed request headers, what architectural considerations would you make to keep the gateway performant and the client code maintainable?
Follow-up Questions
- Can you sketch the exact code you’d write for the prepareHeaders function?
- What edge cases might cause the header injection to fail at runtime?
- How would you verify that the correct headers are being sent in integration tests?