shareglint

n8n community node for Shareglint — render Open Graph and social images from HTML templates

Package Information

Released: this month
Downloads: 0 weekly / 0 monthly
Latest Version: 1.0.0
Author: Shareglint

Documentation

n8n-nodes-shareglint

This is an n8n community node. It lets you use Shareglint in your n8n workflows.

Shareglint renders Open Graph and social images from HTML templates: you send a template slug and the values for its fields, and get back the URL of a finished PNG — usually in about a second.

n8n is a fair-code licensed workflow automation platform.

Installation
Operations
Credentials
Compatibility
Usage
Example workflow
Resources
Publishing

Installation

Follow the installation guide in the n8n community nodes documentation. In short: Settings → Community Nodes → Install and enter n8n-nodes-shareglint.

Operations

Operation API call When to use it
Generate Image POST /v1/images?wait=true One event at a time — waits for the finished image URL, usually about a second
Queue Image POST /v1/images Large batches — returns the render ID right away; collect the result with a webhook or Get Render
Get Render GET /v1/images/{id} Check the status of a render requested earlier
Screenshot → Take POST /v1/screenshots?wait=true Screenshot of a public page (resource Screenshot): format, device, viewport, full page (2 renders), dark mode, delay, wait-for selector, hidden selectors, ad blocking
List Templates GET /v1/templates List the templates on your account (built-in and your own)
Batch → Create POST /v1/batches Bulk generation (resource Batch): one image per row of a Google Sheet, a CSV or a JSON array, rendered in the background; returns the queued batch
Batch → Get GET /v1/batches/{id} Status, counts, ZIP URL and per-row images of a batch
Create Signed URL none (GET /v1/account once per run) A signed og:image URL for static sites and CMS fields — nothing renders until the first visit

Generate Image and Queue Image take:

  • Template — picked from the templates on your account, or a slug set with an expression (for example blog-og).
  • Template Fields — name/value pairs for the template's fields, e.g. title, author, site, date, photo for blog-og.
  • Options — Format (png by default, jpeg, webp or pdf — one page at the template size with vector text), Quality (1–100 for JPEG and WebP, default 85), Scale (1× or 2× for raster formats, default 2× = 2400×1260 for a 1200×630 template), Idempotency Key, Template Text Language (en, de, es, fr, pl; the language for text the template adds itself, such as the date or reading time) and Webhook URL (we call it when the render finishes). The output carries format, and the file URL ends in .png, .jpg, .webp or .pdf.

Create Signed URL takes the template, the template fields (add a field named locale for the template text language) and a File Type (.png, .jpg, .webp or .pdf). It reads your account ID and signing secret with your API key and signs /i/{account}/{template}.{ext} exactly as the docs describe; the output is { url, extension }. The first request to the URL renders the image and uses one render of quota; later requests are free redirects.

Batch → Create (bulk generation) takes:

  • Template — the same template list as Generate Image.
  • Rows From — Google Sheets or CSV Link (the default: a sheet shared with Anyone with the link; an …/edit#gid=N link reads that tab; the first row holds the column names), CSV Text (commas or semicolons) or Rows (JSON) (an array of objects with the template field values plus optional id, returned with the row's result, and filename, the file name in the ZIP).
  • Batch Options — Sizes (og, square, portrait, story, pinterest, youtube; one image per row per size, default the template's own size), Format, Quality, Scale, Template Text Language, Batch Name, Column Mapping (template field → column name; without it columns are matched to fields by name, case and spaces ignored), Webhook URL and Idempotency Key.

Every row × size is one render of quota, and the whole batch must fit the remaining quota up front (batch_quota_exceeded, HTTP 402, with needed and available; nothing starts). The node returns the queued Batch at once — id, status (queued, processing, completed), total, counts (pending, rendering, completed, failed), progress, zip (status, url) — and does not wait: rows render in the background for minutes. Continue the workflow from the webhook: set Webhook URL to the production URL of an n8n Webhook node, which receives one batch.completed event (id, status, counts, zip_url) when every row and the ZIP are done; or run Batch → Get later. Batch → Get returns the batch with items (row, id, filename, size, status, render_id, url, error) a page at a time: Items per Page (up to 1000; the API's default is 100) and Items Cursor (the next_cursor of the previous page). Sharing errors come back with their code, for example batch_sheet_not_shared or batch_unknown_column.

The node is also available as a tool for the n8n AI Agent.

Credentials

  1. Sign up at shareglint.com — the free plan includes 100 images a month.
  2. In the dashboard, create an API key. It starts with sk_live_.
  3. In n8n, create a Shareglint API credential and paste the key into API Key.

Base URL defaults to https://shareglint.com/v1. Change it only if you run your own Shareglint instance.

When you save the credential, n8n checks the key against GET /v1/account/usage, so a mistyped key shows up right away rather than on the first workflow run.

Compatibility

Built with @n8n/node-cli 0.53 and tested on n8n 2.44.6 (installed as a community package, credential test and an end-to-end run of the example workflow). The Batch operations are checked with lint, build and a stubbed request run, not yet in a running n8n.

Usage

  • Set Idempotency Key to the post ID or URL. The same key with the same body returns the same response, without a second render and without a second deduction from your quota — re-running a workflow costs nothing.
  • Errors. Each API error has a stable code in error (for example quota_exceeded, sync_overloaded, idempotency_key_reuse) and a human-readable message. The node raises them as n8n errors with the API message; with Continue On Fail enabled the error goes to the output item instead of stopping the workflow.
  • sync_overloaded (HTTP 503). If Generate Image runs out of time, the render is not lost: the error response carries its id. Use Get Render with that ID instead of rendering again, or use Queue Image for large batches.
  • Quota (HTTP 402). Rendering the same template with the same values returns the existing image and does not use quota.

Example workflow

examples/new-post-og-image.json: RSS Feed Trigger (new blog post) → Shareglint · Generate Image (blog-og template with the post's title, author, domain and date; the post GUID as the idempotency key) → Edit Fields with post_url, post_title, og_image_url and render_id.

To use it: in n8n open Workflows → Import from File, choose the file, set your feed URL in the first node, select your Shareglint credential in the second, and connect the last node to WordPress, Webflow, Airtable, Slack or wherever the image URL should go.

Output of the last node for one post:

{
  "post_url": "https://yourblog.com/sleep",
  "post_title": "Ten things that ruin your sleep",
  "og_image_url": "https://shareglint.com/storage/renders/img_01m08s3ez1zcv28h335g9923bw.png",
  "render_id": "img_01m08s3ez1zcv28h335g9923bw"
}

Resources

Publishing

This package lives in a monorepo (integrations/n8n). n8n requires verified community nodes to be published from GitHub Actions with npm provenance (from May 1, 2026). The workflow for that is in .github-workflow-example/publish-n8n-node.yml; GitHub runs workflows only from the repository root, so it must be copied to .github/workflows/publish-n8n-node.yml at the root of the monorepo. The file's header lists the one-time npm setup and the release steps (bump version, push a tag n8n-v<version>).

Local development:

npm install      # Node.js 24 or later
npm run lint     # n8n's community node lint rules
npm run build
npm run dev      # starts n8n with this node loaded

License

MIT

Discussion