BotBuster | AI Detection & Encoding Toolkit

🔍

Conversation Analyzer

Paste a conversation to analyze for AI patterns

Their Messages (paste the suspected bot's messages)

Context (optional - your messages for context)

📊

Detection Results

AI probability assessment

--

Bot Probability

Paste a conversation and click analyze to get results

💬

Craft Test Response

Generate a follow-up message with embedded detection techniques

Testing Intensity

Subtle Direct 3

1 = Natural conversation with subtle tests over multiple messages
5 = Direct bot-detection questions in single message

Techniques to Include

Behavioral Instruction Zero-Width Format Trigger

Generated Response (copy this to send)

Click "Generate Test Message" to create a response with embedded detection techniques...

🎯

Score Their Reply

Paste the bot's answer to your test and rate how "AI" it reads

Their reply to your test message

📋

Campaigns

Track a target across many messages

New campaign name

Platform (optional)

Campaigns live only in this browser. Back them up to move devices or guard against cache-clearing.

🗂️

Active Campaign

Message log, technique coverage & next step

Select or create a campaign to begin.

🔐

Universal Encoder Suite

40+ encoding methods from practical to absurd

Input Text

Output

Select an encoder and enter text...

📟 Classic Encodings

🔑 Substitution Ciphers

🎮 Fun & Silly

👁️ Steganographic / Hidden

🧬 Advanced / Esoteric

🦗

Locust — Layered Cipher Chain

Stack encoders into one absurd pipeline with a shareable recipe

Input Text

Output

Add steps, then Run Forward...

Add a step

Recipe (shareable)

🔒

Encode Hidden Message

Embed invisible text using zero-width characters

Visible Message (what humans see)

Hidden Payload (embedded invisibly)

Encoded Output (copy this entire text)

Encoded message will appear here...

🔓

Decode / Analyze

Detect and extract hidden zero-width content

Paste Suspicious Message

📚

Injection Payload Library

Pre-built payloads for bot detection

📱

Platform Compatibility Guide

Where zero-width injection works (and doesn't)

⚠️ Note: Platform behavior changes frequently. These are based on community reports and may be outdated. Always test on your target platform first!

✅ High Confidence - Works

⚠️ Medium Confidence - Test First

❌ Low Confidence - Likely Blocked

❓ Unknown - Need Testing

📝

Report Platform Behavior

Help improve this guide with your findings

Platform Name

Status

Notes (optional)

🌐 Community Reports

Reports are stored in this browser and merged into the matrix above. Share them as a JSON bundle, or import a bundle/URL from someone else.

🎯

Bot Detection Techniques

Proven methods to identify AI chatbots

⚙️

Ollama Configuration

Connect to a local Ollama instance or Ollama Cloud (ollama.com)

Ollama API Endpoint Local: http://localhost:11434 | Cloud: https://ollama.com

API Key Required for Ollama Cloud. Leave blank for local. Stored only in your browser.

Model Loaded live from your Ollama instance. Click ↻ after changing endpoint/key.

Custom Model Overrides the dropdown when filled. Use for cloud tags (e.g. gpt-oss:120b-cloud) not shown above.

Not connected

🎨

Detection Preferences

Configure how test messages are generated

Gradual Testing Mode Spread detection techniques across multiple messages

Include Zero-Width Injection Embed hidden prompts in generated messages

Aggressive Formatting Tests Include requests that trigger AI formatting patterns

Messages Before Direct Test How many natural messages before obvious tests

2

💾

Data Management

Export and manage your saved data

🛡️

Privacy

What this site does (and never does) with your data

Your inputs — messages, encoded text, campaign notes — stay in your browser and are never sent anywhere. If this site is hosted online, anonymous usage events (which features get used, country, referrer) may be collected to improve it — never the text you type. Opt out anytime.

Anonymous usage analytics Events only, never your content. Honors Do Not Track.