Configuration - ReyGrid Docs

import { ReyGrid } from "reygrid";

const client = new ReyGrid({
  apiKey: "your-api-key",
  baseUrl: "https://reygrid.com/api/v1",
  timeoutMs: 60_000,
  headers: { "X-My-Header": "value" },
});

All options are optional.

Options (`ReyGridConfig`)

Option	Type	Default	Description
`apiKey`	`string`	—	Sent as `Authorization: Bearer <key>`
`baseUrl`	`string`	`https://reygrid.com/api/v1`	API base URL. Trailing slashes are removed
`headers`	`Record<string, string>`	`{}`	Extra default headers for every request
`timeoutMs`	`number`	`60000`	Request timeout in milliseconds
`fetch`	`typeof fetch`	`globalThis.fetch`	Custom fetch implementation
`onError`	`(error, context: RequestContext) => void`	—	Hook called for failed requests

RequestContext is { method: string; url: string; authenticated: boolean }. The SDK sets Accept and Content-Type to application/json unless you override them in headers. The constructor throws if no fetch is available (Node.js earlier than 18 without a fetch option).

Custom base URL

const client = new ReyGrid({
  baseUrl: "https://custom-gateway.example.com/reygrid",
  apiKey: "your-api-key",
});

Custom `fetch`

const client = new ReyGrid({
  apiKey: "your-api-key",
  fetch: (url, init) => {
    console.log(init?.method ?? "GET", String(url));
    return fetch(url, init);
  },
});

Logging errors with `onError`

const client = new ReyGrid({
  apiKey: "your-api-key",
  onError: (error, { method, url }) => console.error(`${method} ${url}`, error),
});

onError is called for network failures and API errors from regular requests. It is not called for errors raised while streaming with streamChat().

Timeouts

timeoutMs aborts the request when it elapses. For streamChat() it covers the entire stream, so raise it for long responses.

Low-level requests

client.request<T>(path, init?) is public and sends an authenticated request relative to baseUrl. Prefer the typed methods.

const res = await client.request<{ success: boolean }>("/ping");