@wriven-ai/client

Official client for the Wriven headless CMS Content Delivery API. Isomorphic (Node 18+, browsers, edge), typed, zero-dependency.

npm i @wriven-ai/client

import { createClient } from '@wriven-ai/client';

const wriven = createClient({
  projectId: 'YOUR_PROJECT_ID',          // dashboard → API Keys
  token: process.env.WRIVEN_TOKEN!,      // wrk_live_… (published) or wrk_preview_… (drafts)
});

// One entry by slug
const post = await wriven.getEntry('blog_post', 'hello-world', { include: 1 });

// A filtered, sorted, paginated list
const { items, total } = await wriven.getEntries('blog_post', {
  filter: { category: 'news' },
  sort: '-publishedAt',
  limit: 10,
});

Typed entries

interface Post {
  title: string;
  body: unknown;            // ProseMirror JSON — render with @wriven-ai/react
  cover: { url: string } | null;
}

const post = await wriven.getEntry<Post>('blog_post', 'hello-world');
post.data.title; // string

Options

| Option | Default | Notes | |--------|---------|-------| | projectId | — | required | | token | — | required; scope decides published vs. drafts | | baseUrl | https://api.wriven.com | self-hosted / regional override | | fetch | global fetch | inject for older runtimes / tests | | timeoutMs | 10000 | per-request timeout | | retries | 2 | retries on network/5xx (reads only) |

Query options

select · filter[key] · sort (- for desc) · page · limit · include (0–3, expands references). cache / next pass through to the framework fetch (e.g. Next.js ISR). Errors throw a typed WrivenError (status, code).

Preview / drafts

Use a wrk_preview_… token — the API returns drafts automatically and marks responses uncacheable.

MIT