Vinder Singh
Back to all work

Live agent, free to try

LinkedIn Thought Leadership Post

A lead editor briefs a team of specialists and decides who works next, until there's a post worth putting your name to: a real point of view, sourced, and in your voice.

Try it live Say who you are and what to write about, or leave the topic out.
My role
Designed the editor, the six specialists, their briefs and the review standards
The team
A lead editor with a trend scout, researcher, angle strategist, ghostwriter, critic and finisher
Made as
A lite version of an agent I designed at gimmefy
Try it
Live version
The live LinkedIn Thought Leadership Post agent: a box for who you are and what to write about, the editor and six specialists listed on the left, and two example runs Live · Click to try
The public version, live now. Paste a past post or two and it matches your voice.

The problem

You can spot an AI-written LinkedIn post in one line. Confident and vague, a made-up anecdote, a statistic with no source, and a question at the end begging for comments.

The posts worth reading do the opposite. They take a position, back it with evidence, deal with the best argument against it, and admit what the author doesn't know.

How it works

It isn't a fixed production line. The editor reads the brief and decides who works next, the way a real editor runs a desk.

  1. Trend scoutOnly called when you haven't picked a topic. Finds three timely topics from the last two weeks of news in your field.
  2. ResearcherFacts, numbers and named examples, each sourced and dated, plus the strongest counter-argument.
  3. Angle strategistThree different points of view you could take, and what each one risks. You pick one.
  4. GhostwriterWrites the post in your voice, using only what you told it.
  5. CriticPasses the draft or sends it back with up to five specific fixes.
  6. FinisherThe final post, two alternative opening lines, a first comment with the sources, and hashtags.

Key decisions

An editor, not a pipeline

A brief with a topic skips the trend scout. A weak draft goes back to the ghostwriter. The route is decided on every run, so nobody does work the post doesn't need.

The counter-argument is part of the research

The researcher has to bring the best case against your view. A post that has already dealt with the obvious objection is far harder to dismiss in the comments.

No invented experiences

The ghostwriter can only use what you gave it. If a line would need a story you never told, it doesn't get written.

The critic has standards, not opinions

Every fact sourced and dated, no engagement bait, nothing left in that only makes sense to the author. In one example run it caught a vague source date, a personal claim the author never made, and a bait question at the end. All three were fixed before the author saw a draft.

What it produced

7agents, one in charge
3angles to choose from
5fixes max per review

Read both runs, start to finish: a marketer new to agentic workflows, with no topic given and a project manager on meetings people don't dread.