LinkedIn Profile Comments
/linkedin-profile-commentsReturns the comments a LinkedIn profile left on other people's posts — their opinions verbatim, with replies and parent-post context, for 0.2 credits.
Overview
linkedin-profile-comments calls the SyncGTM MCP `linkedin_profile_comments` tool to return the comments a person has left on other people's LinkedIn posts, each with its replies and the parent post's author and engagement counts. It is the strongest personalization signal of the three profile-activity tools, because it gives you the prospect's opinion in their own words instead of a like. Cost is 0.2 credits per call, flat, whether you pull 3 comments or 50.
What it does
- 1
Takes one LinkedIn profile per call as `profile_url` — the full URL or a bare username works
- 2
Returns each comment's text and `posted_at` date, its replies, and the parent post's author plus reaction and comment counts, so you read the opinion in context
- 3
`max_results` accepts 1–50 and defaults to 3 — the skill raises it to 25, since a wider sample costs the same 0.2 credits
- 4
`posted_limit` narrows by recency: `any`, `24h`, `week`, `month`, `3months`, `6months`, defaulting to `any`
- 5
Does not return the person's own posts, emails, phones, job history or education — those are `linkedin_profile_posts` (0.3), `find_work_email` (1) and `linkedin_profile_enrich` (1)
- 6
Costs 0.2 credits per call, flat, and the skill quotes the total spend before it starts a batch
How to use it
- 1
Download the skill file — it saves as linkedin-profile-comments.zip.
- 2
In Claude, open Settings → Capabilities → Skills and upload the .zip (or unzip the linkedin-profile-comments/ folder into .claude/skills/ for Claude Code).
- 3
Connect the SyncGTM MCP server — browser sign-in, no API key — then paste in the LinkedIn profile URLs or usernames you want read.
- 4
Type /linkedin-profile-comments with the profiles and the job: positions argued, an outreach opener, objection prep, or a fallback check on a silent profile.
- 5
Tune `posted_limit` — `week` or `month` for live outreach timing, `3months` or `6months` to prove a consistent point of view. Skip `24h`; it often returns nothing, and the wider retry is a second call at another 0.2 credits.
Use cases
Cold openers that quote
Pull 25 comments before writing a first line, then open by engaging with a position the prospect actually argued. The opener quotes their wording instead of complimenting their profile.
Silent-profile fallback
Many people comment far more often than they post. When `linkedin_profile_posts` comes back thin, run this before writing the prospect off as inactive.
Pre-call objection prep
Read what an attendee has argued for and against in the last quarter, and turn it into the two objections to expect on the call. Their public disagreements predict the ones you will hear.
Warm-thread mapping
See which authors and threads a prospect keeps showing up under. Those repeat authors are your intro paths and the communities worth posting in.
GTM workflow examples
Monday opener pass on this week's outbound list
- 1Source the week's targets with `find_people` — `current_titles` of "VP Sales", `person_countries` of "United States", `current_company_headcount_ranges` of `201-500`, `limit` of 25 — and keep the LinkedIn URLs (0.3 per result, 7.5 credits).
- 2Run `linkedin_profile_comments` on each URL with `max_results: 25` and `posted_limit: month` (0.2 each) to collect comment text plus every parent post's author.
- 3For anyone who returns nothing, fall back to `linkedin_profile_reactions` (0.2) so a like on a relevant post still gives you a hook.
- 4Get the sending address with `find_work_email` (1 credit each) and confirm deliverability with `verify_email` (0.3 each).
- 5Write one two-line opener per prospect that argues with or extends a position they wrote, quoting their exact sentence.
Outcome: Up to 25 contacts with a verified work email, each carrying a first line built on a sentence the prospect wrote themselves, for roughly 45 credits.
Pre-call brief for tomorrow's discovery meeting
- 1Run `linkedin_profile_comments` on each attendee with `max_results: 25` and `posted_limit: 3months` (0.2 each) to capture their positions and the threads they joined.
- 2Run `linkedin_profile_posts` (0.3 each) on the same profiles to separate what they publish officially from what they argue in other people's threads.
- 3Run `linkedin_profile_enrich` (1 credit each) for work experience and education, so you know whether a position dates from this job or the last one.
- 4Assemble a one-pager per attendee: quoted positions, the vendors and categories they named by name, and the two objections those comments predict.
Outcome: A one-page pre-call brief per attendee with their quoted positions and the two objections to expect, for about 1.5 credits a head.
Turn a competitor's thread into a qualified warm list
- 1Run `linkedin_post_commenters` on a competitor or category post with `sort_order: Most relevant` — one page per call at 0.3 credits, so pull page 1 for the top commenters.
- 2Run `linkedin_profile_comments` on the top 20 of them with `max_results: 25` and `posted_limit: 3months` (4 credits total) to see whether that comment was a one-off or a standing position.
- 3Drop anyone whose comment history reads as a vendor pitching or a job seeker, keeping the 8–10 who repeatedly argue about the problem you solve.
- 4Run `find_work_email` (1 credit each) on the survivors and sequence them with an opener that references the thread plus the position they have repeated.
Outcome: A 10-person warm list where every contact has argued about your problem in public within the last quarter, built for under 20 credits.
Prompts
Paste any of these into Claude once the skill is installed and the SyncGTM MCP is connected.
Show me the comments linkedin.com/in/janedoe has left in the last month. Set max_results to 25 and posted_limit to month, and quote each comment verbatim next to the parent post's author.Pull 50 comments from linkedin.com/in/janedoe with posted_limit 6months, then list the positions she has actually argued — what she is for and what she is against — with her exact wording beside each one.Run linkedin_profile_comments with max_results 25 and posted_limit 3months on these 12 profiles: <paste 12 linkedin.com/in/ URLs, one per line>. Return one table: name, most recent comment date, position argued, and the best quote to open with.linkedin.com/in/janedoe has no recent posts. Check her comments instead with max_results 25 and posted_limit 3months, and if that is empty too, check her reactions before calling her inactive.Run find_people with current_titles "VP Sales", person_countries "United States", industries "Software Development", current_company_headcount_ranges 201-500 and limit 20. Then run linkedin_profile_comments on each result with max_results 25 and posted_limit 3months, and draft a two-line opener per person that engages with one position they argued. No compliments.Best practices
Raise `max_results` off its default of 3. Three comments is a spot check; 25 shows a consistent point of view and the call costs the same 0.2 credits.
Set `posted_limit` deliberately. `week` or `month` for outreach timing, `3months` or `6months` for a pattern — `24h` often returns nothing, and re-running wider bills the profile a second time, 0.4 for one person.
Name the tool in the prompt or say "comments they left". Unnamed, your client may reach for `linkedin_profile_posts` (0.3 credits) or `linkedin_profile_reactions` (0.2 credits) instead.
Ask for quotes, not a summary. "Quote her wording" gives you material you can respond to; a paraphrase is something you have to trust.
Never parrot the comment back at the prospect. Engaging with the position works — repeating what they said reads as a bot.
Use this as the fallback for a quiet profile, and stop before the wrong tool: identity and history are `linkedin_profile_enrich` (1 credit), and everyone who commented on a post you already have is `linkedin_post_commenters` (0.3 credits per page, `page_number` 1–10). Run `check_credits` (free) before a batch.