Use Manus To Write Better Expert LinkedIn Posts That Actually Sound Like You

Start a new project in Manus and tell it what you want to build. Give it links to your public writing samples, then ask it to interview you so it can learn your expertise and write better posts in your voice.

You do not need a giant system prompt. Use a prompt like this to get the project moving:

Help me build a simple content workspace that turns my work experience into LinkedIn posts in my voice.

Start by reviewing these public writing samples: [LINKS]. Tell me which pages and files you can and cannot access. Then interview me one question at a time so you can learn my expertise, opinions, and natural writing style. Keep everything inside this project, and do not publish anything.

Manus can use your public posts to spot recurring topics, phrases, opinions, and patterns in how you explain things. It will combine that context with the answers you give during the interview.

Pro tip: Review and disable unused connectors first. Manus may choose an enabled tool automatically. During our test, it started using Notion, so we stopped the run and corrected the destination.

Manus will interview you with questions designed to uncover stories hiding inside your normal workday. It might ask what changed your mind recently, what worked much better or worse than expected, or what changed how you now do a real task.

Here is an adapted version of the first question we used:

Think about the tools, workflows, or decisions you tested in the last 7–14 days.

Tell me about one moment that changed your mind—something that worked much better or worse than expected, exposed a misleading popular belief, or changed how you now do a real task.

As you ramble, include whatever you remember: what you were trying to accomplish, which tools were involved, what happened on the first attempt, what happened after iteration, what surprised or annoyed you, and who should care. If several examples come to mind, share all of them.

Dictate a brain dump in response. Do not worry about structure, grammar, or whether the story is ready to become a post. Talk through what you expected, what actually happened, what surprised or annoyed you, and why someone else might care.

Answer any focused follow-up questions. If a story depends on a result, audience response, or number Manus cannot access, give it the source or flag that detail for verification.

This surfaced two useful stories in our test. One was about a basic one-on-one workflow that became more popular than much flashier AI topics. The other came from an older Microsoft Copilot recording we abandoned when the tool struggled with basic Excel formulas and formatting. The product may have improved since then, but the broader problem remains: people still ask for help with software their employers require, even when an earlier test went badly.

Neither idea started as a polished content pitch. They came out of talking through a real incident.

Once you finish answering, Manus automatically pulls out your strongest ideas, the supporting evidence available in your answers and project, and exact phrases that sound like you. It saves promising stories as idea cards and uses the interview to build a voice and expertise profile.

In this setup, you do not need to ask for each file separately. The opening project request tells Manus to standardize the process, so it can turn the messy interview into the supporting files while it works.

The idea card shown below came from the older Copilot story. It captured the failed recording, who still needs help, and the boundary that product-specific advice would require a current retest. Manus also saved phrases from the dictation instead of smoothing everything into generic AI language.

That last part matters. The rough edges are often where your voice lives. If you use a phrase you would actually say, keep it. If Manus turns it into something that sounds like corporate thought leadership, flag it during review.

Manus also turns patterns across the interview into supporting files, such as content pillars and future interview exercises. Over time, the project becomes a library of your ideas instead of a folder full of disconnected drafts.

Manus does this work on its own computer, building the files while the task runs.

Pro tip: Let Manus finish building the workspace before taking over its computer. Taking control interrupts the flow, so leave the task running in the background and check back later.

Manus automatically turns the strongest idea cards into LinkedIn drafts. In our test, it created two pilot posts from the stories in the interview without needing a separate drafting prompt.

Your job is to review the drafts and give simple, specific feedback:

• Which phrases sound exactly like you?
• Which lines feel too polished or generic?
• What important detail or opinion is missing?
• Is Manus stating anything more confidently than the evidence supports?

For example, you can respond naturally:

The opening sounds like me, but the middle is too polished. Keep my original phrase about [PHRASE], add the missing detail about [DETAIL], and make the conclusion more direct. Use this feedback to write version two and update my voice profile.

Each review gives Manus a clearer record of your preferences. You are showing it where the first draft captured your voice, where it lost it, and what to change in version two.

Pro tip: Review and approve version two before posting. Check what sounds like you, what feels too polished, what is missing, and whether the evidence supports the draft.