How to Auto-Draft LinkedIn and X Posts From a New Blog Post With n8n

The generic version of this automation, new blog post triggers an instant, unreviewed post to every social channel, breaks in practice for two reasons that have nothing to do with n8n's own capability. Posting automatically to a company's official LinkedIn page requires clearing an API approval process most small teams never complete, and since February 2026, X charges per post created under its API, with a materially higher rate for any post containing a link, which a blog cross-post always does. Here's a workflow that handles both realities honestly instead of assuming either platform still works the way it did a few years ago.

For background on n8n's self-hosted pricing model and how it compares to Zapier or Make, see our full n8n profile.

Step 1: Trigger the Workflow From an RSS Feed or CMS Webhook

Start the workflow with n8n's RSS Feed Trigger node pointed at the blog's published feed, or a webhook triggered directly by the CMS on publish if the CMS supports outbound webhooks, which fires faster and more reliably than waiting for an RSS poll interval. Either trigger gives the workflow the new post's title, URL, and summary content the moment it's published, which is everything the rest of the workflow needs to start drafting from.

Step 2: Fetch the Full Article Text, Not Just the RSS Summary

Add an HTTP Request node to pull the full published article by its URL, since most RSS feeds only include a truncated summary or excerpt, which isn't enough context for the next step to draft genuinely accurate, platform-native social copy. Extracting the full body text here, rather than working from a thin summary, is what separates a social post that actually reflects the article's real content from one that reads as a generic, could-be-about-anything teaser.

Step 3: Generate Platform-Specific Drafts With an AI Node, Not One Shared Caption

Use an AI node (n8n supports OpenAI, Anthropic, and other model providers natively) to generate two distinct drafts from the extracted article: a LinkedIn-style post with a professional framing and a natural paragraph structure, and a shorter X post or thread that fits that platform's more compressed format and tone. Generating genuinely separate drafts per platform, rather than one caption reused everywhere, is what makes the output usable without a full manual rewrite, which is the actual point of automating this step in the first place.

Step 4: Route Both Drafts to a Human Approval Step Before Anything Publishes

Send both AI-generated drafts to a Slack channel, an email, or a lightweight approval table (NocoDB or Google Sheets both work well for this) where a real person reviews and approves or edits each draft before it goes anywhere near a publish action. This step exists for two concrete reasons, not just general caution: an AI-drafted post can misrepresent the article or use an off-brand tone, and once X charges per post created, an unreviewed automated posting error is now a real, avoidable cost every time it happens, not just a reputational risk.

Step 5: Understand What LinkedIn's API Actually Requires Before Building the Publish Step

Before wiring up automatic publishing to a company's official LinkedIn page, confirm the account meets LinkedIn's actual requirements: posting on behalf of an organization page runs through LinkedIn's Community Management API, which requires a registered legal entity (LLC, corporation, or nonprofit, not an individual or unregistered side project), a verified LinkedIn Page with a super admin who approves the application, and a two-tier review process that includes a screencast demonstrating the exact use case. This is a genuinely heavier process than generating an API key, and it's worth confirming the organization can actually clear it before designing the workflow around automatic company-page posting. A simpler, immediately available alternative is posting to an individual's personal profile using the more accessible member-level posting scope, which is a reasonable default for a smaller team, with the tradeoff that the post appears under a person's name rather than the brand page.

Step 6: Budget for X's Per-Post Pricing Before Assuming It's Free

As of February 2026, X's default API pricing for new developers is pay-per-use: roughly $0.015 per post created, rising to about $0.20 per post if it contains a link. Since a blog cross-post is a link by definition, budget for the higher rate on every single post this workflow publishes to X, and treat that as a real, ongoing operating cost tied directly to publishing volume, not a one-time setup consideration. For a team publishing a modest number of posts a month this cost is likely trivial in absolute terms, but it's worth confirming explicitly rather than assuming the API is still free, since building on that assumption is exactly the mistake this kind of pricing shift catches teams on.

Step 7: Add a Dedup Check Before Any Publish Action

Log every article URL the workflow has already processed, in the same Google Sheets or database table used for approval tracking, and check incoming triggers against that log before drafting or publishing anything new. RSS feeds and webhook retries can occasionally fire more than once for the same item, and without a dedup check, a workflow that costs real money per X post can silently double-post and double-bill for a single article, which is a failure mode worth designing out from the start rather than discovering after the fact.

Step 8: Test With a Real, Low-Stakes Post Before Trusting the Full Loop

Publish one genuinely low-stakes blog post through the complete workflow, trigger, extraction, AI drafting, approval routing, and the actual publish step, before relying on it for a real announcement or launch post. Confirm the approval notification actually reaches the right person, that edits made during approval are correctly reflected in what publishes, and that the dedup log is updated properly, since any gap here is far cheaper to catch on a test post than on content the business actually cares about getting right.

Why Self-Hosting n8n Specifically Matters for This Workflow

This particular recipe is a reasonable case for running n8n's self-hosted Community Edition rather than a managed, per-task-billed platform, worth calling out explicitly since the choice isn't automatic. The workflow runs on a schedule or webhook trigger with a handful of steps per article, article extraction, AI drafting, approval routing, dedup logging, publish, which on a metered platform accrues a per-task cost on top of whatever LinkedIn and X themselves already charge. Running the orchestration layer itself on a self-hosted instance removes that second layer of metering entirely, so the only real per-action costs left are the ones the destination platforms charge directly, which makes it easier to reason about the workflow's true operating cost without an extra billing layer stacked on top of it.

A Few Mistakes That Undermine This Workflow

A handful of habits reliably undo an otherwise well-built version of this automation. Skipping the human approval step to make the workflow feel more "automated" removes the one safeguard against an AI-drafted post that misrepresents the article or gets published incorrectly, at a moment when a publishing mistake on X now carries a direct dollar cost, not just a delete-and-apologize fix. Assuming a company LinkedIn page can be posted to automatically without first confirming the account has cleared LinkedIn's actual API approval requirements means discovering that gate only after the rest of the workflow is already built. And treating X's per-post pricing as negligible without actually calculating it against real expected publishing volume risks a genuine, avoidable surprise on a bill once the automation scales past the volume it was originally tested at.

What This Workflow Actually Produces

Built this way, full-article context feeding genuinely platform-specific drafts, a real human approval gate, and an honest accounting of what each platform actually requires and costs, this setup delivers what the generic version of this automation only promises: a fast, mostly-automated path from publishing a blog post to genuinely usable social drafts, without pretending either platform still offers frictionless, free, unreviewed posting. That's a meaningfully more durable foundation than a workflow built on assumptions about API access and pricing that were already out of date by the time it went live.

If your team is deciding between n8n's self-hosted control and a fully managed platform like Zapier or Make for this kind of workflow, our Zapier vs Make comparison covers that tradeoff from the managed-platform side.