Let your AI assistant edit your site, safely.

One MCP server over Ghost, WordPress, Shopify and Webflow. Every publish is previewed, version-locked and reversible.

Install

Connected in three steps

1Create a workspace API key

Create API key

Why

The key begins with sk_specter_, is shown once, and is scoped to that one workspace. It authenticates MCP and REST alike, and is revocable from the screen that made it.
2Paste the config into your assistant
Run it in a terminal.
terminal

claude mcp add --transport http spectersync https://app.spectersync.com/api/mcp \ --header "Authorization: Bearer sk_specter_YOUR_KEY"
Why

Registers the server for the current project. Add --scope user to register it for every project instead.
Edit the config file, then restart the app.
claude_desktop_config.json

{ "mcpServers": { "spectersync": { "command": "npx", "args": [ "-y", "mcp-remote", "https://app.spectersync.com/api/mcp", "--header", "Authorization:${SPECTERSYNC_KEY}" ], "env": { "SPECTERSYNC_KEY": "Bearer sk_specter_YOUR_KEY" } } } }
Why

The file is at ~/Library/Application Support/Claude/ on macOS and %APPDATA%\Claude\ on Windows. It takes local commands rather than URLs, so the hosted endpoint is reached through the mcp-remote bridge — which needs Node 18 or later, because that is what supplies npx.
Put it in ~/.cursor/mcp.json.
.cursor/mcp.json

{ "mcpServers": { "spectersync": { "url": "https://app.spectersync.com/api/mcp", "headers": { "Authorization": "Bearer sk_specter_YOUR_KEY" } } } }
Why

That covers every project. For one project only, use .cursor/mcp.json in the repo root instead.
Open Cascade's MCP panel and edit the file it points at.
mcp_config.json

{ "mcpServers": { "spectersync": { "serverUrl": "https://app.spectersync.com/api/mcp", "headers": { "Authorization": "Bearer sk_specter_YOUR_KEY" } } } }
Why

The file is ~/.codeium/windsurf/mcp_config.json; refresh the panel afterwards so Cascade reconnects. Windsurf's docs take serverUrl or url, and their own remote example uses serverUrl.
Put it in ~/.gemini/settings.json.
settings.json

{ "mcpServers": { "spectersync": { "httpUrl": "https://app.spectersync.com/api/mcp", "headers": { "Authorization": "Bearer sk_specter_YOUR_KEY" } } } }
Why

Or .gemini/settings.json in a repo root for that project only. Restart the CLI and run /mcp to see the tools it picked up. The streamable-HTTP key is httpUrl — url is the SSE one.
Put it in ~/.gemini/config/mcp_config.json.
mcp_config.json

{ "mcpServers": { "spectersync": { "serverUrl": "https://app.spectersync.com/api/mcp", "headers": { "Authorization": "Bearer sk_specter_YOUR_KEY" } } } }
Why

Or .agents/mcp_config.json for one workspace. Antigravity reads serverUrl only — its docs say url and httpUrl are not supported, so a config copied from another editor will fail here.
Settings → Connections → MCP servers → Add a custom MCP.
the three fields

Server URL https://app.spectersync.com/api/mcp Header key Authorization Header value Bearer sk_specter_YOUR_KEY
Why

Devin takes a remote server through a form rather than a file. Set the transport type to HTTP (the docs call it Streamable HTTP in prose), paste the three values below, and set Authentication method to Auth Header. Adding a custom server needs the Manage MCP Servers permission; without it the button is not there, and an org admin has to grant it — or you can use Suggest MCP Integration to ask.
Put it in .vscode/mcp.json.
.vscode/mcp.json

{ "servers": { "spectersync": { "type": "http", "url": "https://app.spectersync.com/api/mcp", "headers": { "Authorization": "Bearer ${input:spectersync-key}" } } }, "inputs": [ { "type": "promptString", "id": "spectersync-key", "description": "Spectersync API key (sk_specter_…)", "password": true } ] }
Why

Or run MCP: Open User Configuration from the command palette to apply it everywhere. The inputs block makes VS Code prompt for the key once and store it securely, so a config you commit never carries the key itself.
Export the key, then run it in a terminal.
terminal

codex mcp add spectersync \ --url https://app.spectersync.com/api/mcp \ --bearer-token-env-var SPECTERSYNC_KEY
Why

export SPECTERSYNC_KEY=sk_specter_YOUR_KEY first. The command writes an [mcp_servers.spectersync] block into ~/.codex/config.toml with url and bearer_token_env_var, and Codex builds the Authorization: Bearer … header from that variable at connect time, so the key never lands in the file. Codex's MCP docs cover the same server under config.toml; the flags come from the CLI's help.
ChatGPT cannot reach Spectersync today. No config

Why

ChatGPT speaks MCP through a custom connector, and a connector will not take a bearer key: it authenticates over OAuth, over no authentication, or a mix of the two, per OpenAI's developer-mode guide. A key is what Spectersync issues, so there is no configuration here that would work if you pasted it. The route that needs no OAuth is a Custom GPT with an Action over the REST surface, and that surface is not live either. Both are being built.
3Ask it something read-only
prompt
```
List my connected sites in Spectersync. Do not change anything.
```
Why

Restart the client first so it picks the server up. The prompt should call list_connections and come back with the CMSes on your workspace. A key that is wrong or revoked returns 401 instead, which tells you the same thing faster than a failed write would.

When you ask it to change something, it drafts into a working copy and then calls create_review, which hands you one link with a card per change: before, after, and accept, reject or skip. Nothing reaches your CMS until you accept, and the link expires within a day. That is the whole publish path on the free plan — publish with confirm: true, publishing without the round trip through you, is part of Pro.

Pricing

What a run costs

Reading, previews, rollbacks and the review link cost nothing. Publishing costs 1 credit per document; recipes 1, or 2 for analysis. Signing up grants 400 credits, once, and Pro replaces the meter with a fair-use ceiling.

Free & packs

$0, or $29 a pack

No card, no subscription

400 at signup, $29 per 400 after

Get started free

Pro

$99/mo

Billed monthly, cancel anytime

Unlimited, fair use 5,000 docs/mo

Get started free

Enterprise

Contact for quote

Set to your volume

Contact sales

Plan comparison
Feature	Free & packs	Pro	Enterprise
Snapshot on publish, and one-call revert
Snapshot history kept	7 days	90 days	90 days
AI recipes, single and bulk	—
Analysis recipes	—
Site index — edits that know the rest of your content	—
AI credits	400 at signup, $29 per 400 after	Unlimited, fair use 5,000 docs/mo	Set to your volume
Additional users	—	$49/mo each	Included in the quote
Support	Email	Email	Onboarding + named contact

Free and credit packs share one feature set: the safe path. Pro adds recipes, bulk and the site index. All plan detail →

The workspace

What the key also opens

Every site in one place, as markdown.

post.md

---
title: Winter service intervals
slug: winter-service-intervals
status: published
tags: [maintenance, seasonal]
---

It reads the archive before it edits a page.

Recipes for the passes you would do by hand.

meta_description

- Service intervals | Blog
+ When to book a winter service, and what
+   a workshop checks at each interval.

Four platforms, one set of moves.

Pull to markdown ✓ ✓ ✓ Beta

Publish back ✓ ✓ ✓ Beta

Products — ✓ — —

Draft & published ✓ ✓ ✓ Beta

Full matrix

FAQ

Before you connect it

Can it publish something I have not seen?

No. On the free plan it cannot publish from the chat at all: it stages the edits and sends you one link, and only what you accept there goes live. On Pro, where it can publish directly, a publish takes two calls — a preview that reaches no CMS and returns a single-use id, then a publish carrying that id. A change staged afterwards voids it.

How does the free plan publish?

Through one link. Your assistant stages the edits and asks for a review; you get a page with a card per change — before, after, the checks we ran — and accept, reject with a note, or skip each one. Only an accept reaches your site, and it runs the same snapshot-and-lock path as the paid route. The link expires within a day and takes the staged edits with it, so silence publishes nothing.

What if someone edits the page in the CMS meanwhile?

The write is refused with a 409 and nothing changes. On Shopify and Webflow, whose APIs have no such lock, Spectersync re-reads the live item and compares first.

How do I undo a publish?

Every publish is snapshotted first and returns a snapshotId; one call puts the page back. If the snapshot cannot be saved, the publish does not happen.

What can it change?

Ten fields: title, slug, status, tags, excerpt, feature image, author, body, meta title and meta description. Not themes, plugins, members, customers, orders or newsletters.

Where do my CMS credentials go?

Into the workspace, encrypted at rest with AES-256-GCM under a server-held key, and never into the conversation. Your assistant sees markdown, frontmatter and diffs.

Who builds Spectersync?

Axel Antas-Bergkvist, under Brilliant Rebels in Sweden, on the Shopify, WordPress and Ghost sites he runs himself. More on the about page.