Automated Content Generator:
X and LinkedIn

Posting consistently was a full-time job on top of the actual work.

The client had built a strong audience across X and LinkedIn but was burning more than 10 hours every week just on content. Not on strategy or building a body of work. On the manual loop: research a topic, draft a post, find or create an image, format it for each platform, schedule it, and repeat. Every week. Without fail.

The problem was not motivation. The problem was that consistency at scale requires infrastructure, not discipline. No person can maintain a high publishing cadence on two platforms indefinitely while also running a business. Something always breaks first, and it is usually the content calendar.

The ask was direct: build a system that produces and publishes platform-native content on both X and LinkedIn automatically, on a schedule, with zero required human input after setup.

The biggest risk was generic output. AI content that sounds like AI content destroys credibility faster than silence.

Before writing a single node, I worked with the client to define what good content actually looked like for their brand. I pulled their top 20 performing posts from both platforms and identified the patterns: the hook structures that performed, the post lengths that got engagement, the topic categories their audience responded to, and the tone that felt like them.

That analysis became the prompt architecture. Rather than giving Claude a generic "write a LinkedIn post about AI" instruction, I built structured prompts that embedded the client's voice, enforced platform-specific formatting constraints, and required the AI to ground each post in a real, verifiable trend rather than producing vague observations.

Platform differentiation was also non-negotiable. LinkedIn and X are different products with different audiences and different norms. A LinkedIn post that performs is long-form, structured, and professionally framed. An X post that performs is punchy, opinionated, and formatted for the feed. The system writes for each platform independently rather than cross-posting the same content.

A fully autonomous publishing pipeline. From trending topic to live post with no human in the loop.

• 01
  Trend research

  The workflow fires on a scheduled trigger and pulls trending topics from multiple sources relevant to the client's niche. It filters for recency and relevance, then selects the best-fit topic for the day's content. This ensures every post is grounded in something real rather than recycled evergreen content.
• 02
  Platform-specific content generation

  Claude AI receives the selected topic alongside the client's voice reference, platform formatting rules, and post-type instructions. It produces two separate pieces of content: a structured LinkedIn long-form post and a short-form X post. Each is reviewed by a validation node that checks format, length, and tone before proceeding.
• 03
  Custom image generation

  Flux AI generates a custom image for each post based on a prompt derived from the content. The image is styled to match the client's visual brand and sized correctly for each platform. This is what separates the output from generic AI content: every post looks visually intentional, not stock-photo generic.
• 04
  Automated publishing

  The content and image are posted to LinkedIn via the LinkedIn API and to X via the X API at the optimal posting times for each platform. A log entry is written to a Google Sheet so the client can review the week's output in one place without needing to check each platform individually.

The client publishes more consistently now than when they were doing it manually.

The most immediate change was that the content calendar stopped being a source of anxiety. The client no longer thinks about what to post. The system handles research, writing, imagery, and publishing. They check the Google Sheet log when they want to, but they do not have to.

The second change was consistency. When content creation is manual, life interrupts it. A busy week means no posts. With this system, a busy week looks identical to a quiet one from the audience's perspective. The publishing cadence became completely independent of the client's schedule.

The custom image generation was the element that exceeded expectations. The client had expected AI-generated images to look off. The Flux AI output, styled against their visual reference, looked intentional. Several followers asked who their designer was.