Turn one post into many outputs

A good blog post is raw material for a dozen other things — a newsletter issue, a social thread, a LinkedIn write-up, a short summary for a roundup. The work of repurposing is rarely the hard thinking; you already did that when you wrote the post. It is the mechanical lift of getting the text in front of a tool that can reshape it, one post at a time, over and over. When your whole archive is a folder of files, that lift disappears.

Why files change the math

Inside Ghost, your posts are locked behind the web editor — you cannot easily feed them to an AI assistant in bulk, so repurposing stays a copy-paste-per-post ritual. Specter syncs your blog down to local markdown, which means every post is a plain text file an AI tool can read directly. Point Claude, ChatGPT, Gemini, or another assistant at the folder and the entire back catalog becomes source material it can work through at once. Specter bundles no AI of its own; you bring the model, and Specter just makes your posts reachable. Set it up with the first-sync guide if you have not.

What a repurposing pass looks like

Because the outputs usually live somewhere other than your blog — your email tool, a social scheduler — this is mostly a read-and-generate job rather than a write-back-to-Ghost one. You ask the AI to read posts and produce new artifacts:

Read this post and draft a newsletter issue from it: a subject line,
a short personal intro, the key points in my voice, and a link back
to the full article. Keep it tighter than the original.

Or work across the archive to find what is worth resurfacing:

Look through every post in this folder. Suggest ten that would make
strong standalone social threads, and for the top three, draft a
thread of 5–7 posts each, faithful to the original argument.

The posts stay exactly as they are on Ghost — you are spinning out derivatives, not editing the source. If a repurposing pass does inspire an edit to the original (a clearer intro, a fixed fact), that change syncs back through Specter the normal way, with a dry-run preview before it goes live.

Keep your voice in it

The reason to do this with your own posts and your own AI, rather than a generic content tool, is that the source is you — your argument, your phrasing, your evidence. Prompt for fidelity to the original and review what comes back, the same human-in-the-loop discipline as editing posts with Claude. Repurposing should amplify what you already said well, not flatten it into the same shape every other account is posting.

This is one of a family of AI jobs that open up once the blog is files: refreshing old posts, translating the whole blog, and more. The common thread is that your archive stops being a closed CMS and becomes a corpus your tools can actually use.