Add Sprites to GitHub Copilot
Available on Copilot in VS Code, and Copilot CLI. Any Copilot plan that includes agent mode. You need a Sprites account with WordPress already connected — the connector acts on that connection, so GitHub Copilot never handles a WordPress credential.
Open the MCP configuration
Run MCP: Add Server from the Command Palette and choose Workspace or Global. Workspace writes .vscode/mcp.json, which is worth committing so the team shares the same connector; Global keeps it to you across every project.
VS CodeCommand PaletteMCP: Add ServerWorkspaceAdd Sprites as an HTTP server
The top-level key is servers, not mcpServers, and a remote endpoint needs an explicit type of http. This is the shape VS Code and Copilot share.
.vscode/mcp.json{ "servers": { "sprites": { "type": "http", "url": "https://mcp.sprites.ai/mcp" } } }Authenticate
VS Code shows an Auth action above the server entry once the file is saved. Run it, sign in to Sprites in the browser window that opens, and approve the scopes.
Spriteswants access on behalf of Copilot- Read your connected ad accounts and site data
- Draft campaigns, creatives and pages for your approval
- Apply only the changes you approve
AllowCancelSwitch Copilot Chat to agent mode
MCP tools are only offered to the agent. Pick Agent from the mode dropdown in Copilot Chat, then open the tool picker and confirm Sprites is listed.
Agent modeCopilot Chat → mode dropdown
Official WordPress MCP vs Sprites MCP
WordPress.com runs its own MCP server at public-api.wordpress.com/wpcom/v2/mcp/v1. Here is what each one covers, so you can pick the one that fits in GitHub Copilot — or connect both.
WordPress.com hosts its own MCP server, now able to write: 19 abilities across posts, pages, comments, categories, tags and media, each switched on individually in your MCP settings, with the agent describing every change and asking before it applies. Writes need a paid WordPress.com plan. Self-hosted WordPress has no hosted equivalent; its official route is the WordPress/mcp-adapter plugin over the Abilities API in WordPress 6.9.
| WordPress.com officialpublic-api.wordpress.com/wpcom/v2/mcp/v1 | Sprites MCPapproval-gated | |
|---|---|---|
| Which WordPress | WordPress.com sites; self-hosted sites install the MCP Adapter plugin instead | Self-hosted WordPress, over the REST API with an application password or the Sprites plugin |
| What it writes | Posts, pages, comments, categories, tags and media, each toggled on separately | Posts and page edits, with metadata, schema and internal links set |
| Before a change applies | The agent describes it and asks you to confirm in the chat | A draft or a diff you review; going live is a separate approval |
| What it knows about SEO | Your site’s content, analytics and settings | Your site plus Search Console, keyword data and what already ranks |
| Choosing what to write | Left to the model and your prompt | Researched: the keyword, the gap, and the pages that should link to it |
If you run WordPress.com and want to draft or tidy content from a chat, its own server is well built — every write is confirmed and every operation can be switched off. Use Sprites when the question is what to write and why: it reads Search Console and keyword data first, works on self-hosted sites without a plugin, and holds every post and edit as a draft you approve.
Run the first prompt
Ask GitHub Copilot this. It routes to a Sprites door, which reads your WordPress account and answers — no account to name, no confirmation to give first.
- With Sprites, write a post on "wordpress speed optimization" and draft it in WordPress.
- Ask Sprites which WordPress pages lost position in the last 28 days and why.
- Use Sprites to rewrite the meta title and description on our /pricing page.
- Have Sprites add an FAQ section with schema to the three posts that get AI Overview impressions.
- With Sprites, cluster our blog by intent and tell me which posts cannibalise each other.
Open the Copilot Chat tool picker in agent mode; Sprites appears with its tools, each one toggleable.
What GitHub Copilot can do with WordPress
The same doors, whichever client you are in. Marketing data inside the editor where the tracking code lives, so a conversion event can be checked against what GA4 actually recorded before the pull request is opened.
The post lands in WordPress, not in the chat
A draft in your posts list with the title, excerpt, focus keyphrase, meta description and internal links already filled in — not a wall of text you have to paste and reformat.
It has read your site first
Internal links point at pages that exist, and the voice matches what is already published, because Sprites crawled the site before it wrote anything.
Page edits arrive as a diff
You see which block changes before anything touches the live site. No agent rewriting a page you cannot recover.
Metadata and schema written with the page
Titles, descriptions, headings and FAQ schema produced alongside the body rather than bolted on by a plugin afterwards.
Search Console is in the same conversation
Queries you rank 8–20 for, pages that slipped, and the post that would close the gap — asked and answered without changing tools.
Publishing is always a separate yes
Every write, a draft included, is proposed for your approval first, and going live on your domain is its own separate approval.
How approvals work here
GitHub Copilot renders tool results as text, so a WordPress change comes back written out rather than as a card. The approval waits for you in Sprites — open it there and the same run continues. Nothing is applied until you approve it.
What it reads
- Your published posts, pages, categories and internal link structure
- Search Console queries, impressions, clicks and average position
- Keyword volume, difficulty and what currently ranks
- Existing titles, meta descriptions and schema
What it can change
- Posts and pages, created as drafts
- Titles, meta descriptions, focus keyphrases, headings and FAQ schema
- Edits to existing pages, delivered as a reviewable diff
- Publishing is a separate, explicit approval — never implied by a draft
When it does not work in GitHub Copilot
Three things go wrong in this client specifically, and each has a one-line fix.
Sprites is configured but Copilot never uses it.
Chat is not in agent mode. Ask mode does not call MCP tools at all.
The config is ignored.
The top-level key is servers. mcpServers is the shape other clients use and VS Code does not read it.
Teammates do not see the connector.
It was added Globally rather than to the Workspace. Move it to .vscode/mcp.json and commit that file; each person still signs in as themselves.
WordPress and GitHub Copilot: questions
Which URL do I paste into GitHub Copilot?
https://mcp.sprites.ai/mcp. There is one endpoint and every client uses it — every OAuth grant is bound to exactly that address, so a variant of it will not be honoured.
Can GitHub Copilot change my WordPress account without asking?
GitHub Copilot renders tool results as text, so a WordPress change comes back written out rather than as a card. The approval waits for you in Sprites — open it there and the same run continues. Nothing is applied until you approve it.
Do I need a WordPress API key or developer token?
No. WordPress is connected once inside Sprites through WordPress (Application Password or the Sprites plugin), and the connector acts on that connection. GitHub Copilot never sees a credential for it.
Why GitHub Copilot for WordPress specifically?
Marketing data inside the editor where the tracking code lives, so a conversion event can be checked against what GA4 actually recorded before the pull request is opened.
Is there an official WordPress MCP server?
Two, one for each kind of WordPress. WordPress.com hosts an MCP server with write tools for posts, pages, comments, media and taxonomies, each change confirmed by you. Self-hosted sites use the official WordPress/mcp-adapter plugin, which exposes the Abilities API from WordPress 6.9 as MCP tools. Automattic’s earlier wordpress-mcp repository has been archived, so any guide telling you to install that one is out of date.
What is the difference between that and Sprites?
The adapter exposes WordPress to a model. It does not do the research, know what already ranks, read Search Console, or write in your voice — it gives an agent the ability to create a post, and leaves the deciding to whatever is on the other end. Sprites is the other end.
Keep going
The same GitHub Copilot connector reaches every other channel — nothing else to add.
WordPress, from GitHub Copilot
Paste one URL, sign in once, and the next WordPress question you ask GitHub Copilot gets answered from the live account.