X Post Commenters
/x-post-commentersPulls every account that replied to an X post — handle, name, bio and comment text, up to 1000 in one call for 0.4 credits.
Overview
X Post Commenters is a SyncGTM MCP skill that takes an X (Twitter) post URL and returns the accounts that commented on it: username, display name, bio and the exact comment each one wrote. One call covers a whole thread, up to 1000 commenters, for a flat 0.4 credits. It returns X accounts rather than LinkedIn profiles and carries no emails, so reaching an inbox is a separate enrichment step.
What it does
- 1
Takes one required input, `post_url` — the full URL of an X post. A profile URL or a bare handle is not valid input; get post URLs from `x_profile_posts` first.
- 2
`max_results` sets the pull size, 1 to 1000, default 100. The default quietly truncates a busy thread, so set it yourself. `post_url` and `max_results` are the only two parameters the tool accepts.
- 3
Returns four fields per commenter: `username`, `name`, `bio` and the `comment` text verbatim — the reply is the personalization hook.
- 4
The `bio` is the only firmographic signal in the response. There is no job title field and no company field to filter on, so bio-less accounts are rarely prospects.
- 5
Returns no emails, phones or LinkedIn URLs. To reach a work inbox, search `find_people` on the company the bio names, match the returned `name` back to the commenter, then run `find_work_email`.
- 6
Costs 0.4 credits per call, flat, whatever `max_results` you set. Asking for 1000 on a 40-reply thread returns 40 and still costs 0.4.
How to use it
- 1
Download the skill file — it saves as x-post-commenters.zip.
- 2
In Claude, open Settings → Capabilities → Skills and upload the .zip (or unzip the x-post-commenters/ folder into .claude/skills/ for Claude Code).
- 3
Connect the SyncGTM MCP server using the setup guide for your client at docs.syncgtm.com/mcp_server/setup. It is a browser sign-in, no API key.
- 4
Paste an X post URL and run /x-post-commenters. If you only have a handle, run `x_profile_posts` with `max_posts: 50` first and pick the thread with the most replies.
- 5
Tune the one setting that matters: `max_results`. Cost is flat, so use 500–1000 on an active thread and keep 100 only for a quick sample.
Use cases
Warm outbound lists
Pull everyone who replied to a competitor's thread about the problem you solve. Each reply is a stated interest you can open the email with.
Voice-of-customer research
Collect verbatim reply text from several posts in your category and cluster it into the objections, tools and phrases buyers use unprompted.
Founder-led demand capture
Review who replied to the founder's posts each week so the SDR follows up with people who already recognize the name.
Launch and event follow-up
Turn the reply thread under a launch or conference post into a named list of interested accounts instead of a screenshot of engagement.
GTM workflow examples
Competitor thread to booked meetings
- 1Run `x_profile_reactions` on the competitor's full X profile URL with `max_posts: 50` (0.2 credits). It returns each post with likes, retweets, replies and views, so rank by replies and pick the thread about the problem you sell into.
- 2Run `x_profile_posts` on the same account with `max_posts: 50` (0.4 credits) to get that thread's post URL — reactions gives you counts, not URLs.
- 3Run `x_post_commenters` on that post URL with `max_results: 1000` (0.4 credits) to get every replier's handle, name, bio and comment text.
- 4Drop rows with an empty bio, then run `find_people` with `current_company_names` set to the companies those bios name and `limit: 25` (0.3 per result). `find_people` has no person-name filter, so match the returned `name` back to the commenter yourself.
- 5Run `find_work_email` on the exact matches only (1 credit each), then `verify_email` (0.3 each), and write each first line from that person's own reply rather than the competitor's post.
Outcome: A verified, contactable list of people who publicly said they have the problem, each row carrying their own words as the email opener.
Monday review of the founder's own posts
- 1Run `x_profile_posts` on the founder's handle with `max_posts: 20` (0.4 credits) — the default is 10, so set it — to list last week's posts and their URLs.
- 2Run `x_post_commenters` with `max_results: 500` on the two or three posts that drew real discussion (0.4 credits each) to pull the accounts that replied.
- 3Deduplicate by `username` and flag every bio that names a title or a company — that flagged slice is the buyer share of the audience.
- 4Run `find_people` with `current_company_names` and `current_titles` taken from those bios and `limit: 25`, keep only rows whose `name` matches the commenter, then run `find_work_email` (1 credit each) on those LinkedIn URLs.
Outcome: A 10–30 name Monday follow-up queue of people who engaged with the founder by name, with work emails attached and the reply text as context.
Messaging brief from category chatter
- 1Run `x_profile_posts` on five creators in your category with `max_posts: 25` (0.4 credits each, 2 credits total) and shortlist the posts about the problem you solve.
- 2Run `x_post_commenters` on five of those posts with `max_results: 1000` (0.4 each, 2 credits total) to collect several hundred verbatim replies.
- 3Group the `comment` text into themes — problems named, objections raised, tools mentioned, questions asked — and count replies per theme.
- 4Run `linkedin_post_commenters` (0.3 credits) on a LinkedIn post covering the same topic to check the themes hold where profiles carry job titles.
Outcome: A one-page messaging brief with five to six ranked themes, verbatim quotes under each, and the competitor names buyers raise unprompted.
Prompts
Paste any of these into Claude once the skill is installed and the SyncGTM MCP is connected.
Get the commenters on https://x.com/stripe/status/1234567890123456789 with max_results 1000. Return username, name, bio and comment, and sort the rows whose bio names a job title to the top.Pull the last 20 posts from the X profile lennysan, pick the three with the most replies, then get the commenters on each of those posts with max_results 500.Commenters on https://x.com/notionhq/status/1234567890123456789, 300 results. Drop anyone with an empty bio and give me the rest as a table with the comment text kept verbatim.From the commenters on https://x.com/figma/status/1234567890123456789 with max_results 800, keep only bios that name a company, then take the top 20 of those. Run find_people with limit 10 on each of those company names to find the matching LinkedIn profile, and find work emails for exact name matches only. Show me the credit spend before enriching.Get the commenters on https://x.com/hubspot/status/1234567890123456789 and https://x.com/gong_io/status/1234567890123456789, 1000 results each, then group the comment text into themes and tell me which tools and objections come up most.Best practices
Set `max_results` on purpose. Cost is a flat 0.4 credits per call whether you ask for 100 or 1000, so the default of 100 only loses commenters on a busy thread.
Budget by post count, not commenter count. Twenty-five posts is 10 credits; the replies inside each thread are free.
Filter on `bio` before you enrich. `find_people` bills 0.3 per result and `find_work_email` 1 credit per person, roughly 1.3 credits per contactable name — 300 replies enriched raw runs near 400 credits, the 12 whose bio names a role runs about 16.
Pick topic posts over brand posts. Replies to "here is how we forecast" are prospects; replies to a product announcement are mostly existing customers.
Use `x_profile_reactions` (0.2 credits) to find which post drew the discussion. It needs a full `profile_url` and returns per-post likes, retweets, replies and views — counts, never the accounts and never the post URLs, so you still need `x_profile_posts` (0.4 credits) for the URL.
Feed it post URLs only, and run `check_credits` (free) before working through a list of posts. A profile URL or a bare handle is not a valid `post_url`; get real URLs from `x_profile_posts` first.