TikTok Comments From Post
/tiktok-comments-from-postPulls up to 100 comments and commenter profiles from one TikTok video for a flat 0.4 credits — username, comment text, likes and profile URL.
Overview
TikTok Comments From Post calls the SyncGTM MCP `tiktok_comments_from_post` tool to return the comments left on one TikTok video plus the profile behind each comment — username, comment text, like count and profile URL. One call reads one video, up to 100 comments, for a flat 0.4 credits. Use it to mine the language your market actually types before you use it as a lead source: TikTok profiles rarely carry a company or a job title.
What it does
- 1
Takes one required input, `post_url` — the full TikTok video URL. A handle or a profile URL will not work.
- 2
`max_results` sets volume: 1 to 100, default 10. Raise it to 100, because the cost does not move with it.
- 3
Returns four fields per comment: `username`, comment `text`, `likes` and `profile_url`.
- 4
Returns no email, no phone, no job title and no company. Contact data has to come from a separate enrichment step, and that step costs more than the scrape.
- 5
Costs 0.4 credits per call, flat. Video count is what multiplies spend — 20 videos is 8 credits.
- 6
Reads one video per call with no pagination past 100 comments. For more coverage, read more videos.
How to use it
- 1
Download the skill file — it saves as tiktok-comments-from-post.zip.
- 2
In Claude, open Settings → Capabilities → Skills and upload the .zip (or unzip the tiktok-comments-from-post/ folder into .claude/skills/ for Claude Code).
- 3
Connect the SyncGTM MCP server — browser sign-in, no API key — then paste in the TikTok video URL. If you only have a topic or a handle, give it that instead and the skill runs `tiktok_search_query` (0.4 credits) or `tiktok_profile_video` (0.5 credits) first to get URLs.
- 4
Type /tiktok-comments-from-post and say how many videos to read. The skill states the credit cost (videos × 0.4) before it calls anything.
- 5
Tune one setting: `max_results`. It defaults to 10, which is a sample. Set it to 100 — the call bills the same either way.
Use cases
Voice-of-customer mining
Pull 100 comments from the top videos in your category and lift the exact wording buyers type into subject lines, ad hooks and landing page bullets.
Objection library
Count which complaints repeat across comment threads on competitor videos, then write the rebuttal into the sequence before an SDR hears it live.
Warm list from a viral video
Filter commenters down to the ones who name a business or a role, then match only those on LinkedIn before spending a credit on enrichment.
Creator vetting
Read the comments under a creator's best videos to check whether their audience is buyers or browsers before paying for a partnership.
GTM workflow examples
Rewrite the campaign in your buyer's words
- 1Call `tiktok_search_query` with your category keyword and `max_results: 50` (0.4 credits) to get the videos ranking for the topic, and rank them by comment volume.
- 2Call `tiktok_comments_from_post` on the five most-commented videos with `max_results: 100` each — 2 credits for up to 500 comments.
- 3Group the comment text into problem, objection and tool-mention themes with a count per theme, and pull the three highest-count phrases verbatim.
- 4Rewrite the cold email subject line, the first opener line and the top three landing page bullets around those phrases.
- 5Call `check_credits` (free) to log the run cost against the campaign.
Outcome: A one-page voice-of-customer sheet — five ranked buyer phrases with comment counts — already written into a live email opener and landing page copy, for 2.4 credits.
Warm list from a competitor's best-performing video
- 1Call `tiktok_profile_video` on the competitor handle with `max_posts: 30` (0.5 credits) to list recent posts as `description`, `posted_at` and `url`, and pick the video on your category topic.
- 2Call `tiktok_comments_from_post` on that video URL with `max_results: 100` (0.4 credits) to get usernames, comment text, likes and profile URLs.
- 3Keep only commenters whose username or comment names a company, product or role — typically 10 to 20 of the 100 — and drop the rest.
- 4Call `find_people` with `limit: 1` per survivor to match each to a LinkedIn profile (0.3 credits per result), then run `find_work_email` (1 credit) and `verify_email` (0.3 credits) on the matches only.
- 5Load the deliverable addresses into a sequence where line one references the problem they described under the video.
Outcome: A 10–20 name outbound list with deliverable work emails and a per-person hook taken from their own comment, built for about 25 credits instead of the roughly 160 that enriching all 100 commenters would cost.
Pick the TikTok creator worth sponsoring
- 1Call `tiktok_search_query` on your category with `max_results: 50` (0.4 credits) and shortlist the five accounts that appear most often.
- 2Call `tiktok_profile_video` per shortlisted handle with `max_posts: 100` (0.5 credits each, 2.5 total) to read posting cadence and pick each creator's two strongest videos.
- 3Call `tiktok_comments_from_post` on those two videos per creator with `max_results: 100` (0.4 each, 4 credits across ten videos) to see whether the audience asks buying questions or fan questions.
- 4Call `scrape_emails_from_website` on the link-in-bio domain of the top three creators (0.5 credits each, 1.5 total) to get a published contact address.
Outcome: A five-creator shortlist ranked by audience fit — posting cadence plus a buying-question share backed by comment counts — with a contact email for the top three, for 8.4 credits.
Prompts
Paste any of these into Claude once the skill is installed and the SyncGTM MCP is connected.
Search TikTok for "cold email tips" and return the top 25 videos, then pull 100 comments from the video with the most comments and group them into themes.Collect 100 comments from this TikTok video and tell me the five problems people keep describing: [paste TikTok video URL]Get the last 20 TikTok posts from the username duolingo, then pull 100 comments from its two most-commented videos and tell me who that audience actually is.Pull 100 comments from this TikTok video, keep only the commenters whose username or comment names a business or a role, and give me that shortlist with profile URLs: [paste TikTok video URL]Run check_credits, then collect 100 comments each from these 5 TikTok video URLs, report the repeated objections, and stop and tell me credits used: [paste 5 TikTok video URLs]Best practices
Set `max_results` to 100. The default of 10 is a sample and the call costs the same 0.4 credits either way.
Chain a search first. `tiktok_search_query` (0.4 credits) finds the videos worth reading; this tool only reads a URL you already have.
Budget by video, not by comment. Video count is the multiplier — 20 videos is 8 credits — so cap the batch and run `check_credits` (free) before working a long list.
Do not expect firmographics. No company, no title, no email, no phone comes back. Budget the enrichment step separately: `find_people` bills 0.3 credits per result and its `limit` defaults to 25 (7.5 credits), so pass `limit: 1` per person, then `find_work_email` at 1 credit each.
Do not call a found address verified. `find_work_email` finds it; only `verify_email` (0.3 credits) checks that it is deliverable.
When the buyer lives on LinkedIn, use `linkedin_post_commenters` (0.3 credits) instead — those profiles carry job titles, so the list needs no matching step.