Agents In Use
What’s new todayLast update
108 use cases added today

Comet researches ten days of founder posts

Mehta reports that Comet navigated Aravind Srinivas’s profile, expanded recent posts and read comment threads to summarise ten days of Comet-related updates and reactions.

FROM THE SOURCE

Original prompt

Shared by the source author. See source notes for context.

Go to Arvind Srinivas' profile and check what he's posted about Perplexity Comet in the last 10 days.

Evidence at a glance

Sourced account
Source author
Geetika Mehta
Original account
Read the source ↗
Published
Source link checked
Reported result
Mehta reports that Comet navigated Aravind Srinivas’s profile, expanded recent posts and read comment threads to summarise ten days of Comet-related updates and reactions.
Human involvement
Mehta evaluates the returned summary, but the account does not independently audit coverage or validate the underlying company claims.

A linked account supports attribution. It does not mean we independently reproduced its result.

Report an issue with this caseOpen correction email

Opens your email app with the case link included. Nothing is sent automatically.

How the agent did it

  1. Name the founder profile, subject and ten-day research window.
  2. Navigate the profile and expand relevant recent posts.
  3. Inspect comment threads for reactions and concerns.

Geetika Mehta gives Comet a bounded research task: inspect Aravind Srinivas's recent profile posts about the browser over the preceding ten days. The assistant navigates to the profile, expands posts and examines multiple updates, rather than relying only on snippets visible in a search result.

Its summary covers discussions of phishing-email protection, Comet Plus, early-adopter feedback and the Invisible team acquisition. Comet also enters comment threads to capture community reactions and concerns. Those two evidence types serve different purposes: the founder's posts describe the company's perspective, while replies show how some readers responded.

Mehta reports receiving a comprehensive summary within minutes, but does not independently measure coverage or fact accuracy. The output is research into public communications, not verification that every discussed feature or business claim is true. The useful scope is a named person, a named subject and a defined date window, followed by inspection of expanded posts and their replies. A repeatable version should preserve source links and distinguish statements by the author from audience commentary. The suggested prompt here reflects those reviewable outputs; the original prompt excerpt remains the author's actual instruction, including its spelling of the first name.

Prompt
Original prompt

Source: Blog · Geetika Mehta

Frequently asked questions

Why expand posts rather than use snippets?

The reported workflow inspected the full recent updates and their comment threads, preserving more context than a search snippet.

What subjects appeared in the result?

The summary covered phishing-email protection, Comet Plus, early-adopter feedback and the Invisible team acquisition.

Does the summary verify those claims?

No. It summarises the founder’s perspective and community reactions; it is not independent verification of each feature or claim.

More details

Remaining steps

  1. Summarise the founder’s updates separately from community commentary.

About this account

Comet navigates and expands multiple profile posts, follows comment threads and synthesises a bounded research summary across those pages.

Firsthand account by Geetika Mehta; original publication date verified against Primary newsletter byline August 31, 2025. Outcomes are the author’s reports, not independently reproduced. Original profile-research prompt shared by the author; spelling preserved.

Credit: Geetika Mehta

Source published

Added to Agents In Use

Source checked

Tools and topics

Perplexity Comet · Founder profile · Comment threads

Date-bounded research · Founder updates · Community reactions

Editorial completeness: 100/100. Measures documented fields, not agent performance or verified outcomes.