Use case

Read LinkedIn posts on any topic with your AI agent

Ask your agent who posts about your topic on LinkedIn and which posts get replies. A real run: the prompt, the call and the posts that came back.

Jesse Sibley
By , FounderPublished · Last updated · 4 min read

At a glance

JobWho posts on a topic, and which posts get replies
Toolslinkedin_search_posts, then linkedin_get_comments
Credits1 for the search in this run
NeedsNo LinkedIn login, cookie or API app
Run on"AI search visibility", past month, 26 October 2026

The prompt

We asked our own agent, Claude Code connected to Manifold, on 26 October 2026:

TEXT
Search LinkedIn for posts about AI search visibility from the past month. Rank them by likes
and comments, tell me how big each author's following is, and say what each post asks the
reader to do.

The call

StepToolInputCredits
1linkedin_search_posts"AI search visibility", past month1

One page of LinkedIn's ranked results: 10 posts, each with its text, likes, comments, the author's follower count and up to 10 of its top comments. The next page, or one post's full list of commenters with linkedin_get_comments, is 1 credit more.

What came back

10 posts, with 2,577 likes and 1,216 comments in all. Here they are by likes, each with its opening line:

The dates LinkedIn shows are ages, such as "2 weeks ago", so each post's date is exact only to that unit.

What the agent concluded

  • A few authors hold the reach. Three posts, from two authors with over 200,000 followers each, took 2,508 of the 2,577 likes (97%) and 1,200 of the 1,216 comments (99%). No other post had more than 28 likes.
  • Most posts sell something. Six of the ten send the reader to the author's own course, report, guide, scorecard or product, or to a tool. One of the most liked recommends Searchable, an AI visibility tool whose LinkedIn ads on AI visibility our ad library search found the same day.
  • The talk is in the comments. On the two posts with over 450 comments, the top replies discuss how to measure AI visibility and why mentions on other sites matter. That is where someone with an answer can join in.

Run it on yours

Put your own topic and window in the same prompt:

TEXT
Search LinkedIn for posts about {your topic} from the past {week, month or year}. Rank them by
likes and comments, tell me how big each author's following is, and say what each post asks
the reader to do.

Then ask for the full commenter list on the post you want to join (linkedin_get_comments), or read a competitor's own company posts (linkedin_get_company_posts). Your agent can draft a comment in the chat; you post it yourself.

Where to go next

Questions people ask

Do I need to connect my LinkedIn account?

No. Manifold never signs in to LinkedIn. It reads public posts, profiles and company pages through a data provider, so there is no login or cookie of yours to share.

Does the search return every post?

No. LinkedIn search is ranked and returns one page at a time. Search two or three phrasings of the topic, and read a competitor's own posts from its company page.

Can I see who commented?

Yes. Each post comes with up to 10 of its top comments and the commenters' profile links, and the commenters tool returns more for one post, at 1 credit.

How exact are the post dates?

Approximate. LinkedIn shows an age such as 2 weeks ago, so a date is exact only to that unit.

Can my agent post or comment for me?

No. Manifold looks things up. Your agent can draft a comment in the chat, and you post it yourself.

Jesse Sibley

Jesse Sibley

Founder of Manifold. A growth engineer who has worked with London startups for eight years.

  • 4 min read

    Find SEO quick wins in your own Search Console data

    Ask your agent for the queries where you almost rank. A real run on our own Search Console: the prompt, the calls, the output and what it changed.

  • 4 min read

    Build a B2B lead list with your AI agent

    Ask your agent for companies by what they do, where they are and their size. A real run: the prompt, the call, the list it returned and the rows to check.