您的健康状况如何 评论区？
粘贴一个公开的个人资料链接。在30秒内，我们将扫描最后500条评论，并给您一个0-100的健康评分、标记示例和风险分析。
仅限公开个人资料。每天5次免费扫描。无需账户。
四个维度的 评论健康
每次扫描都进行与FeedGuardians生产系统对数百万条评论进行的相同检查。
Spam Detection
Copy-paste bot comments, giveaway scams, affiliate link spam, emoji-only replies, and coordinated follower farms.
Toxicity Analysis
Profanity, hate speech, targeted harassment, dogpiling, coded slurs, and slurs in non-English languages.
Sentiment Mix
Ratio of positive, neutral, and negative tone. Tracks sentiment drift across your last 30 posts.
Response Gap
Customer questions that went unanswered and how long they sat before someone replied — or if no one did.
常见 asked
The tool analyzes the most recent 500 public comments on your profile across your latest posts. It looks at four dimensions: spam rate (copy-paste, bots, links), toxicity rate (profanity, hate speech, targeted attacks), sentiment (positive vs negative tone), and response gap (questions your team has not replied to). You get a 0–100 health score and a list of the top issues.
Yes, the summary is free and requires no signup. To get the full PDF report with flagged comment examples, post-by-post breakdown, and a 30-day monitoring trial, you enter your email. We do not share your email or sell it to third parties — it is used to send the report and (optionally) a weekly health digest.
No. The tool only works on public profiles because it reads the same comments that any visitor to your page can see. If your account is private, connect it directly to FeedGuardians via the official platform OAuth to run a full audit on private comments, DMs, and ad comments.
The classifier uses the same AI models as the FeedGuardians production system, which is benchmarked at 95%+ accuracy on English comments and 91–94% on Spanish, Portuguese, French, and German. Rare languages and coded slurs have lower recall. The tool flags borderline cases rather than hiding them silently, so you can judge for yourself.
Native platform filters only block an explicit keyword list that you manually maintain. They cannot detect sarcasm, coded language, brand-specific attacks, or the kind of "polite" negative comments that erode trust in your ads. FeedGuardians detects context — it knows the difference between "this looks amazing" and "this looks amazing... if you want to get scammed."
Yes, as long as the profile is public. Running it on competitors is a useful way to benchmark your own numbers. Just drop their URL into the input. The free tier runs up to 5 scans per day per user.