Ad filter
The erid marker, hashtags, up to 10 of your own words and phrases, plus any links, @mentions, phone numbers and emails. If a rule fires, the post goes to no target of the project.
Decide which posts go to each platform and in what form. Filter out ads at the entrance, and for each target keep only the topics you need, remove foreign links, replace words and trim the footer.
Crosslybot filters work in two stages. At the entrance, in the source settings, everything that should go nowhere is filtered out: sponsored integrations by the erid marker, hashtags, your own words or contacts, and in groups — messages not from admins and replies. Then each target decides which posts it needs: by keywords and hashtags, by content type or by share of posts.
The second stage cleans the text for the platform: link rules remove or replace foreign addresses, word replacement fixes words and emoji, and footer trimming removes the source channel’s signature block. Everything that did not pass the filters is shown in Operations with the reason.
In the source settings — the ad filter for channels, and for groups — admin messages only and skipping replies.
A keyword and hashtag filter, content type (media only, photos and videos, photos only or text only) and the share of forwarded posts.
Link rules, word replacement and footer trimming — separately for each target.
Posts that did not pass the filters do not disappear silently: Operations shows which rule fired.
The erid marker, hashtags, up to 10 of your own words and phrases, plus any links, @mentions, phone numbers and emails. If a rule fires, the post goes to no target of the project.
An allowlist for a target: only posts containing at least one of the words, phrases, hashtags or emoji go there. Up to 50 items, the type is detected automatically.
For each target: all posts, media only, photos and videos only, photos only or text only. Plus a share of posts from 1 to 100%.
Remove a foreign link or replace it with yours — by address or a whole website, separately for each platform.
“Find → replace” pairs for each target: for example, “channel” → “community” or a Premium emoji → a regular one.
Set a marker — for example, “Subscribe” — and the line with it and everything below won't reach the target.
The ad filter is set on the source and filters a post out for the whole project when a rule fires. The keyword filter is set on a target and works the other way round — as an allowlist: only posts with the right words or hashtags go to that target.
A target filter only includes: a post passes if it has at least one word from the list. To keep posts with certain words or hashtags out of all targets, use the source ad filter — it accepts your own words and hashtags.
In the project, Operations → the Filtered tab. It shows the reason: for example, the ad rule that fired or missing keywords. A post filtered by mistake can be published within 24 hours.
No, the post limit is spent only on posts that were actually sent.
The ad filter is off for Telegram and Max groups, but groups have their own settings: admin messages only and skipping replies. More on the groups and forums page.
Those are publishing settings rather than filters — they are on the flexible settings page.
Connect the bot in 2 minutes. Free plan — no cards, no signup.