Everything Sei can ask of a site, what happens when you configure nothing, and how to change it. On WordPress the defaults are fixed; on a custom site each operation is a handler you choose to write.
Check connection (ping)
Sei asks what the site can do, when you connect it and when you click Check connection.
| Default | Change it | |
|---|---|---|
| WordPress | Sei reads the REST API, the plugin's status and the user's capabilities | Install the plugin for SEO fields |
| Custom site | @getsei/webhook answers with the capabilities your handlers give |
onPing adds name, timezone, and drafts or schedule |
Sei stores the answer: a feature appears in Sei only when the site says it can do it. After you add a handler, click Check connection again.
Publish (article.publish)
Create or update an article: live now, as a draft, or at a scheduled time.
| Default | Change it | |
|---|---|---|
| WordPress | Gutenberg blocks; SEO fields with the plugin | Nothing to set |
| Custom site | onPublish (required) stores what it wants |
Read contentHtml or contentBlocks; store the cover from featuredImage; refuse a mode with 422 unsupported-mode |
| Payload | Lexical rich text, SEO fields in plugin-seo's meta or a sei group |
fields, seo, featuredImage, taxonomies options; EXPERIMENTAL_TableFeature on the content field keeps tables (without it, one paragraph per row) |
Conflicts: send back a revision and Sei sends it as ifRevision next time. Throw SeiConflict(current) when it
no longer matches, and the writer chooses Overwrite, Pull into editor or Save as new draft.
Unpublish (article.unpublish)
Take a live article back to a draft.
| Default | Change it | |
|---|---|---|
| WordPress | Status back to draft | Nothing to set |
| Custom site | Not offered | Add onUnpublish |
Import (article.list)
Bring the site's existing articles into Sei, 100 at a time.
| Default | Change it | |
|---|---|---|
| WordPress | Every post Sei can read, with categories, tags and featured image | Nothing to set |
| Custom site | Not offered | Add onList; report featuredImage, categories and tags to bring those too |
Articles in Sei's Trash are skipped, so an import never brings back what you deleted.
Categories and tags (terms.list)
| Default | Change it | |
|---|---|---|
| WordPress | The site's categories and tags; a new one is created after you confirm | Nothing to set |
| Custom site | Sei sends categories and tags with every publish; storing them is up to you |
onTerms lists yours in Sei's pickers; capabilities: ["taxonomies"] if you store them without listing |
| Payload | Off | taxonomies: true, or name your own collections |
Cover image (featuredImage, 1.3)
| Default | Change it | |
|---|---|---|
| WordPress | Uploaded to the media library once, set as the featured image | Nothing to set |
| Custom site | article.featuredImage = { url, alt } comes with every publish that has one |
Store the URL, or copy the file into your own storage |
| Payload | Not stored | fields.featuredImage: a text field (URL) or an upload field (downloaded once) |
Removing the cover in Sei leaves the site's image as it is.
Slug changes
Changing the slug of a live article changes its URL. Links to the old URL, and the old URL in search results, keep working only if the site redirects it. Sei warns in the editor before you publish a new slug.
| Default | Change it | |
|---|---|---|
| WordPress post | WordPress redirects the old URL to the new one by itself (it keeps the old slug) | Nothing to set |
| WordPress page | WordPress doesn't keep old slugs for pages; the GetSei plugin (1.2 or newer) does, and redirects them with a 301 | Keep the plugin active |
| Custom site | Nothing: the old URL answers 404 | When onPublish updates an article whose slug changed, keep the old slug and answer it with a 301 to the new URL |
| Payload | @getsei/payload 1.5 keeps the old slugs in sei.previousSlugs |
Call findSeiRedirect from your page route when a slug finds nothing, and redirect (its README has the snippet) |
On a custom site the receiver already knows the old slug: onPublish gets the article by externalId before it
saves the new one. The demo blog keeps every slug a post has had and redirects them, for example in Next.js:
// app/[slug]/page.tsx
const post = await getPost(slug);
if (!post) {
const moved = await findByPreviousSlug(slug);
if (moved) permanentRedirect(`/${moved.slug}`);
notFound();
}Delete (article.delete, 1.2)
Delete in Sei always moves the article to Sei's Trash for 30 days. With Also delete it on your site ticked:
| Default | Change it | |
|---|---|---|
| WordPress | The post goes to WordPress's Trash | Nothing to set |
| Custom site | Not offered: the checkbox doesn't show | Add onDelete: delete for good and answer recoverable: false, or hide it and answer recoverable: true |
| Payload | Deleted for good | delete: { field: "deletedAt" } keeps it; delete: false never deletes |
Restore (article.restore, 1.2)
Restore from Sei's Trash brings the article back in Sei. On the site:
| Default | Change it | |
|---|---|---|
| WordPress | The post leaves WordPress's Trash as a draft | Nothing to set |
| Custom site | Only in Sei; a later publish creates a new article | Add onRestore for articles onDelete kept |
| Payload | Only with delete: { field } |
Clears the field, back to draft |
Preview (article.preview, 1.4)
Preview opens the draft in a new tab, in the site's own look. On the site:
| Default | Change it | |
|---|---|---|
| WordPress | Plugin 1.2 or newer: your theme renders the draft in place of a published post | Nothing to set |
| Custom site | The draft is placed into one of your published articles' pages | Add onPreview: your own article page renders it, exactly |
| Payload | The draft is placed into one of your published articles' pages | Coming to @getsei/payload |
Coming next
The same rule applies to what's next: a default that works, and a hook to change it. Planned: onMedia (images in
the body uploaded to your storage instead of Sei's) and a beforePublish transform.