You know you should post on LinkedIn. You haven't in weeks. The activation energy is absurd: you shipped three features this week but you can't make yourself write 200 words about any of them.

So you avoid it. Or you paste from ChatGPT and hope nobody notices. Either way, your professional network has no idea what you're building, and every week of silence is compounding opportunity cost you can feel but can't quantify.

The problem was never discipline. It's that the platform asks you to perform when you'd rather be performing in the technical sense.

LinWheel is an MCP server that gives your coding agent a full LinkedIn publishing pipeline. Draft, reshape through seven content angles, generate branded cover images, build carousel PDFs, schedule, publish. One npm package.

You keep building. When you have something worth sharing, you tell your agent. It writes, renders, and queues. You approve or reject from your terminal. That's your entire LinkedIn workflow now.

Voice profiles learn from your approvals and rejections. After a few runs through the loop, the drafts stop reading like they were written by three LLMs in a trench coat and more like something you'd actually pen. No other tool on the market learns your voice from usage. Your craft is the input; a social presence is the output.

MCP is the inflection point. Agents can now discover and use tools without custom integrations. Every developer who uses a coding agent already has the infrastructure to publish. They just don't have the tools.

The window is open: the people who build with agents are the same people who hate performing on LinkedIn. They're our users. And they're adopting MCP tooling right now.

Peleke Sengstacke is a technical writer (Scotch.io top 15 globally, first full-time Curriculum Engineer at Trilogy Education [acq. $750M]) and an engineer who builds production AI systems, including qortex, a framework that lets agents actually learn from their mistakes and get measurably better at tasks over time. He also has ADHD and would rather mass-delete his LinkedIn than mass-produce content for it.

LinWheel exists because the founder needed it. The adaptive voice learning is powered by qortex, the same learning engine behind his other products. The visual engine uses Satori + resvg-wasm. Zero external image APIs. Zero per-image cost. From idea to published post without leaving your terminal.

This is founder-market fit: someone who understands the problem because he lives it, knows what bad AI content looks like because he's a writer, and built the infrastructure to close the gap.