Read and post to Tumblr blogs, manage likes and follows, and check a user's dashboard, over the Tumblr API v2.
Tumblr ships in the w6w first-party pack. It declares 23 actions, 1 health check, and the host runs its code in a sandbox that never sees the credential.
io.w6w.tumblrTumblr is a blogging and social-networking platform built around Blogs and Posts. This app covers a blog’s public surface (info, avatar, likes, following/followers) and, once connected, its private surfaces too: published, queued, draft and submission posts; creating, fetching, editing and deleting a post in the Neue Post Format (NPF); and a post’s notes (likes, reblogs, replies).
It also covers the connected user’s own account: profile and blog list, per-feature rate-limit headroom, the dashboard feed, liked posts, followed blogs, and the four one-shot actions (follow, unfollow, like, unlike) that make a workflow act like a Tumblr user rather than just read one.
Authorization is OAuth 2.0 — Tumblr’s newer, now-recommended protocol, alongside the OAuth 1.0a it has supported since launch (see the README for why OAuth 2.0 was chosen and where the two auth levels the API actually documents, ‘API key’ and ‘OAuth’, line up with it). Health checks track Automattic’s own shared status page, which names Tumblr’s API, Dashboard and Sites as three of its ~30 monitored components.
Three routes to the same 23 actions. The Workflow tab is generated from Tumblr's own manifest and carries its real ids, so it is copy-pasteable; the Code and CLI examples are the same call for any action on any app, so every app-specific value in them is a blank you fill in.
blog-avatar-get Fetch the URL of a blog's avatar image, in one of 9 supported sizes.
blog-following-list List the blogs this blog follows, most recently-followed first.
blog-likes-list List the posts a blog has publicly liked. Likes sharing must be enabled on the blog.
blog-posts-submission-list List posts submitted to the blog by other users, awaiting review.
user-dashboard-get Fetch the connected account's dashboard feed. Do not use this to rebuild a Tumblr client — see the vendor's API usage policy.
A workflow step names the app and the action, and the editor fills in the
connection when you pick one. This is the Step shape from the
workflow spec, carrying Tumblr's real ids.
{
"manifestVersion": "2",
"name": "tumblr-example",
"steps": [
{
"id": "blog-avatar-get",
"uses": {
"app": "io.w6w.tumblr",
"action": "blog-avatar-get",
"connection": "conn_YOUR_CONNECTION_ID"
},
"with": {
"blogIdentifier": "<blogIdentifier>"
}
}
]
}blog-avatar-get blog-followers-list blog-following-list blog-info-get blog-likes-list +18 more actions available
Every app-specific value here is a blank you have to fill in. An
app action is reached through the connection that authenticates it, so the
address is a connection id, not the app id — and connections belong to your account,
so a public page cannot know yours. Create one for Tumblr, then fill in
the three blanks: conn_YOUR_CONNECTION_ID, the action key, and the
parameters that action declares. The call itself is real — the shape is transcribed
from the studio's own snippet builder, which prints the same kind of blanks — but
nothing in it is specific to Tumblr. The Workflow tab is where this app's
real ids are.
npm install @w6w/sdkyarn add @w6w/sdkpnpm add @w6w/sdkdeno add npm:@w6w/sdkimport { W6wClient, isActionRun } from "@w6w/sdk";
// Reads W6W_BASE_URL and W6W_TOKEN from the environment when omitted.
const client = new W6wClient();
const envelope = await client.run({
urn: "conn_YOUR_CONNECTION_ID",
action: "blog-avatar-get",
payload: {
blogIdentifier: "<value>",
// size: "<value>",
},
});
if (isActionRun(envelope)) console.log(envelope.value); npm install -g @w6w/cli w6w run conn_YOUR_CONNECTION_ID --action blog-avatar-get --payload '{"blogIdentifier":"<value>"}' Give an AI agent Tumblr — without giving it Tumblr's credentials. One MCP endpoint exposes every app, function and workflow the caller is entitled to, as tools it can discover and run. Access is granted per team while we onboard.
One tool call{
"name": "w6w_invoke",
"arguments": {
"ref": "app:io.w6w.tumblr#blog-avatar-get",
"input": {
"blogIdentifier": "<blogIdentifier>"
}
}
}
Every tool names its target with a single ref. The
app: form above doesn't name a connection at all — the
host resolves which of the caller's Tumblr connections to sign
with, and refuses rather than guesses when the answer is ambiguous.
The token is attached host-side, at the moment of the call. It is never a tool argument, never in the model's context, and never in a transcript — so a prompt injection has nothing to exfiltrate.
Tools are derived per end user from what that person has actually connected and is entitled to — not one shared bot identity carrying the union of everyone's access.
Multi-step work runs on the workflow engine and returns a run handle the agent can poll — retries, branching and state survive the conversation that started them.
Tumblr's declared health checks are on the surface too, so an agent can tell "the vendor is down" from "your credential expired" before it burns a retry on either.
The MCP surface is part of the hosted platform. Tumblr itself is MIT, and the runtime that executes it is source-available (FSL).
Tumblr declares its own checks, so its health is a property of the app rather than something the host guesses at.
Reads the three Tumblr-specific components (API, Dashboard, Sites) from Automattic's shared status page at automatticstatus.com, which also covers WordPress.com, Jetpack and ~30 other products.