Specter for Ghost
Connect your Ghost blog in the browser and your whole archive opens as clean, structured content. Run AI recipes across every post at once, review each change as a diff, and Specter publishes back over the Ghost Admin API.
Ghost 5.x · self-hosted or Ghost(Pro) · nothing to install · browsing & diffs are free · workspaces open with 500 free credits.
Also for: WordPress → · Shopify → · Webflow →
Why Ghost users reach for Specter
The Ghost web editor is built for writing a single post. The Admin API is fine for one-off scripts but rate-limits hard on bulk runs. And the "AI for Ghost" plugins on the market are mostly closed-loop GPT wrappers that charge per article and only ever see one post at a time. Specter breaks that ceiling: connect your blog in the browser, run AI recipes with your whole archive as context, review every change as a diff, and publish back over the Admin API — without the timeouts.
What you can do
Specter isn’t an AI and it isn’t a CMS. It’s the workflow and control layer: connect your live Ghost blog in the browser, run AI recipes across the whole archive, and review every change before it ships — so the model can finally reach your content, all of it, at once.
Connect your Ghost blog and every post lands in one workspace as clean, structured content. AI recipes run with your whole archive as context — so they weave accurate internal links, match your tone across the blog, and stop generating duplicates.
Run an AI recipe across thousands of posts — rewrite meta descriptions, fix outdated mentions, regenerate SEO titles, swap CTAs. Review the whole pass as a diff, approve what is right, and publish back over the Ghost Admin API.
Run a translation recipe across every post into five languages, preserving Ghost tags, feature images, and slugs — and review each change before it ships.
AI runs as recipes across your connected content, not pasted post-by-post into a chat window. The whole Ghost blog is the context — and you review every change as a diff before publishing.
Edit in your workspace and the change publishes back. Edit in the Ghost admin and Specter pulls it in. Every post round-trips with frontmatter, tags, and authors preserved.
Every run shows up as a diff first — exactly which Ghost posts would be created, updated, or flagged as conflicts. Browsing and reviewing diffs are free; you approve what ships, and every publish keeps a snapshot for rollback.
If a post changed on both sides since the last sync, Specter pauses and shows you both versions instead of silently overwriting one. If a team edits in the Ghost admin alongside you, this keeps the two sides from clobbering each other.
Connecting Ghost, browsing posts, and reviewing diffs are free. Only AI runs spend credits, and every workspace opens with 500 free — enough to see exactly what a recipe does across your archive. Your content stays yours.
How it works on Ghost
Step
In Ghost: Settings → Advanced → Integrations → Add custom integration. Copy the Admin API key. Walkthrough →
Step
Open your Specter workspace, add a Ghost blog, paste your blog URL + Admin API key. Specter tests the connection, then pulls your whole archive into the workspace. Connect Specter to Ghost →
Step
Edit a post by hand or run an AI recipe across the whole blog. Review every change as a diff, approve what is right, and Specter publishes back to Ghost. Every publish keeps a snapshot for rollback.
What round-trips today
Specter syncs the text content of your Ghost blog — body, frontmatter, tags, authors, scheduled status. Members, newsletters, themes, and Stripe are deliberately out of scope.
Post and page body (projected from Ghost Lexical), title, slug, status (draft/published/scheduled), publish date, tags, authors, custom excerpt, feature image URL and alt, local image upload, codeinjection_head, codeinjection_foot, canonical URL, meta title, meta description, OG title, OG description, Twitter title, Twitter description, visibility (public/members/paid).
Members, subscribers, newsletter sends, Stripe keys, themes, routes.yaml, integrations other than its own, redirects, staff users. Read the full capability matrix → for the exact, current state.
Credits, not lock-in
Connecting Ghost, reviewing diffs, and publishing edits cost nothing. Only AI recipe runs spend credits, and every workspace opens with 500 free.
.md files on diskFAQ
Connect in the browser and start with 500 free credits. Prefer files on your own disk? There’s a desktop and open-source edition too.
Guides
Hand-picked walkthroughs for the things people actually do with Specter on Ghost.
Guide
Add a CTA, affiliate disclosure, or footer to all your blog posts: connect Ghost to Specter in the browser, append the block across the whole archive, review the diff, and publish back.
Read guide →Guide
Bulk edit Ghost tags and clean up a messy taxonomy: connect your blog to Specter in the browser, normalize tags across the whole archive with AI, review the diff, and publish back.
Read guide →Guide
Connect your Ghost blog to Specter in the browser: paste your site URL and Admin API key, test, and run your first sync to pull every post into your workspace.
Read guide →Guide
Keep your blog style guide consistent at scale: connect Ghost to Specter in the browser, apply voice and formatting rules with an AI recipe, review the diff, and publish back.
Read guide →Guide
Update outdated blog content at scale: connect your Ghost posts to Specter in the browser, run an AI recipe to flag stale facts and dates, review the diff, and publish back.
Read guide →Guide
Generate Ghost meta descriptions in bulk with AI: connect your blog to Specter in the browser, fill missing excerpts across the whole archive, review the diff, and publish back.
Read guide →Guide
Bulk Ghost internal linking in the browser: connect your blog to Specter, run an AI recipe to add contextual links and fix broken ones across every post, review the diff, and publish back.
Read guide →Guide
Ghost version history is limited to a few recent edits per post. Specter keeps a snapshot behind every publish, so you can review the diff and revert — or mirror to git for permanent history.
Read guide →Comparisons
Comparison
A Ghost AI writer alternative: instead of autogenerating posts and pushing them live, connect Ghost in your browser, run AI recipes across your existing posts, and review every change as a diff before it publishes.
Read comparison →Comparison
Ghost markdown editing in your own tools vs the browser editor. A fair comparison of when each wins, and how to get local markdown with Ghost.
Read comparison →Comparison
Obsidian Publish turns your vault into a website; Ghost is a full publishing platform with newsletters and memberships. Here's how they compare — and how to keep writing in Obsidian while publishing to Ghost.
Read comparison →Comparison
The Obsidian Ghost Publish plugin pushes notes to Ghost one way. Specter connects your whole Ghost blog in the browser, runs AI recipes across it, and reviews every change as a diff — both directions. Here's the side-by-side.
Read comparison →Comparison
Looking for a Ulysses Ghost alternative or iA Writer Ghost sync that round-trips existing posts? Here is where each app stops and where Specter fits.
Read comparison →Comparison
You can script the Ghost Admin API to sync posts to markdown yourself. Here's the build-vs-buy on what a robust two-way ghost api sync actually takes.
Read comparison →Specter also supports
Platform
Self-hosted WP or WordPress.com Business+, via Application Passwords and the REST API.
Explore WordPress →Platform
Edit your Shopify store blog and articles in markdown. Articles and blogs only — never orders or customers.
Explore Shopify →Platform · Beta
Any Webflow plan with the CMS. OAuth scoped to CMS read/write — the Designer, components, and Ecommerce stay untouched.
Explore Webflow →