moder8.net: An Automated Content Moderation Tool

Last Updated: 2026-08-20

moder8.net is a free Prompt Development & Testing tool for performing automated moderation of user generated content (UGC) using an LLM as the moderation engine.

Instead of a SaaS black box, you develop your own moderation prompts to flag harmful content into one or more of 13 safety categories plus additional brand protection.

Developing Your Own UGC Moderation Prompt

Broadly the process is:

  • Modify the default Preprocessor prompt
  • Modify the default Brand Protection prompt
  • Refine the Safety Categories prompts
  • Test moderation pipeline using Sandbox, Test Bench & Forum Simulator
  • Copy full moderation prompt into your own UGC moderation system

Develop Moderation Prompts

At moder8.net, develop and test your moderation prompt using Preprocessor, Categories, the Sand Box, Test Bench and Simulator.

Notice how the Sand Box, Test Bench and Simulator give you the total number of tokens and the cost of the moderation call as well as a break down into individual cost components.

Modify The Safety Categories

From the Prompts menu, choose Categories. You can see the individual prompt for each safety category. If you make any changes and click Save the changes are merged into the overall content moderation prompt.

You can also restore the default prompt for any given category or insert a 1 shot (example) template.

Check or uncheck the box for a particular category which indicates whether or not it is actively moderated.

moder8 prompts

Manage Brand Protection

Under Prompts > Brand Protection is where you define the LLM prompt that protects your brand against malicious posts and disinformation.

Preprocessor

This is where global prompt logic should go. It is sometimes known as the "goal". Don't waste your time and more importantly, your context window, by assigning the model a personality and giving it a back story.

The Complete Prompt

Select the Prompt menu (not any of the submenus) to display the complete moderation prompt.

moder8 test bench

Once your complete moderation prompt is well tested you can copy it to your own moderation system. Bare in mind it's been developed and tested against gemini-2.5-flash-lite so the results on other models may be different.

Sandbox

The Sandbox allows you to enter or paste in ad-hoc content to test your moderation prompt. When you click Run Moderation you will get a list of any safety categories that have been villated as well as a detailed breakdown of the cost of that particular moderation call.

moder8 test bench

Test Bench

The Test Bench contains 5 adversarial examples for most safety categories

moder8 test bench

Simulator

The Simulator calls the LLM to generate a fictitious forum thread with up to 4 replies. Select the option you want and click Generate Thread. After a brief delay the post and its replies will be displayed.

The original post and the replies may or may not contain violations. Use the Test Pipeline Against Thread button to moderate the OP and replies.

The Simulator is a particularly useful tool for edge case identification.

moder8 Dashboard

Audit Log

The Audit Log retains a complete history of all moderation calls and the result. Click the id of the log entry to see more details.

moder8 Dashboard

Dashboard

The Dashboard displays a summary of moderated UGC and the moderation costs as well as the cost of simulation calls.

moder8 Dashboard

Test Moderation Prompt

Use your LLM of choice to generate new test cases. Below is a sample prompt. 

I am building an automated user content generation (UGC) moderation tool that uses an LLM prompt as the moderation engine to categorise violations into one of thirteen categories in increasing order of severity:

NONSENSICAL, JAILBREAK, SPAM, PROFANITY, SCAM, CRIMINALITY, HARASSMENT, HATE_SPEECH, SEXUAL, VIOLENCE, SELF_HARM, CHILD_SAFETY, PII plus an additional category of BRAND_PROTECTION which is intended to flag content that could damage my brand.

Can you please generate 5 adversarial tests for each category. The difficulty should be as hard as you can make them.

Additionally can you generate some edge cases that touch on one or more of the safety categories.

NEED HELP?

I offer an initial prototype or technical assessment for qualified projects. Contact me  Copy email address for more information.