Introducing CleanSpeak Guard: Real-Time Text Moderation for Bubble Apps, No API Required
Hi everyone,
I just released CleanSpeak Guard, a smart real-time text moderation and filtering plugin for Bubble apps.
CleanSpeak Guard helps detect and clean unsafe, profane, NSFW, insulting, or abusive language before it appears in your app. It works locally inside your Bubble app and does not require an external API, API key, or paid moderation service.
It can be used to censor text, replace unsafe words with safer alternatives, detect unsafe content, or block form submissions through Bubble workflows.
Plugin page:
Demo:
Editor preview:
Key features:
- No API required
- Real-time input checking
- Workflow/server-side text checking
- Light, Medium, and Harsh strictness levels
- Censor mode
- Safe-word replacement mode
- First-letter masking mode
- Custom blocked words
- Custom safe-word replacements
- Allowed words list for exceptions
- Detection-only mode
- Smart fuzzy matching for common bypass attempts
- Handles symbols, spacing tricks, repeated letters, leetspeak, slang, misspellings, and common word combinations
- English moderation support
- Works on web and native mobile
Example use cases:
- Moderating comments before they are posted
- Filtering usernames, bios, chat messages, reviews, or support tickets
- Blocking unsafe form submissions
- Replacing profanity with safer alternatives
- Building safer community, marketplace, education, or social apps
How it works
Add the CleanSpeak Guard element to a page, connect it to your text input, and choose your moderation settings.
You can check text directly, watch an input in real time using a target selector, or run moderation through Bubble workflows. The plugin exposes useful states like:
- Clean Text
- Original Text
- Is Clean
- Match Count
- Matches
For example:
Only when CleanSpeak Guard's Is Clean is yes -> Create a new Comment
Or:
Only when CleanSpeak Guard's Is Clean is no -> Show an alert asking the user to clean up their message
There is also a server-side action for checking or sanitizing text before saving it to your database.
A quick note: no text filter can be perfect in every possible case. CleanSpeak Guard is designed to catch a large majority of unsafe words, slang, misspellings, symbol-based bypass attempts, spacing tricks, repeated-letter tricks, leetspeak, and common combinations, but unusual cases may still slip through. If you find a missed word or bypass pattern, please contact me at bassim3qazi@gmail.com so it can be reviewed for a future update.
Future updates may include optional AI moderation for builders who want deeper context-aware filtering.
Would love to hear feedback, feature requests, and ideas for improving the filter.