Your Knowledge Base
Upload documents to teach your AI new things.
Drag & drop files here or click to browse
Supports PDF, TXT, and MD up to 10MB
Do not upload sensitive documents.
Uploaded Documents
Knowledge Graph Visualization
Explore entities and relationships extracted by LightRAG from your documents.
Click "Refresh Graph" to load or make sure you have uploaded and indexed some documents.
AI Security Hardening
Monitor security events, system shields, and prompt injection logs.
Hermes Adaptive Rules
Self-synthesized regex rules from security log analysis.
Semantic Attack Memory
Vector database cache of unique blocked prompt injection signatures.
| Time | Detected Via | Attack Signature |
|---|---|---|
| No attack signatures cached in semantic memory yet. | ||
System Settings
Configure your AI persona and developer resources.
AI Persona
Define how your AI should behave and respond.
Developer API Keys
Access the OpenRouter-compatible endpoint at /api/chat/completions.
RAG & Knowledge Storage Management
Manage, reset vector databases, or wipe all uploaded knowledge data.
Reset vector databases if changing embedding models, or perform a complete wipe of all documents, vector indices, and knowledge graphs.
API Reference & Tutorial
Learn how to query the completions endpoint.
/api/chat/completions
curl -X POST https://chatbot.kagita.my.id/api/chat/completions \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"message": "Hello, how are you?",
"chatId": 1
}'
Info: chatId (number, optional) links the request to an existing
chat session to retrieve context & save history. Omit for a one-off query.