Chat & RAG puts a personal AI agent inside 📝AI & Augmentation — grounded in your own library via retrieval-augmented generation, able to read, write, and research on your behalf, and configurable down to the model and system prompt. It's for creators who want an AI collaborator that actually knows their knowledge base, not a generic assistant bolted on top.
How It Works
Every MythOS library gets a chat surface — a floating panel reachable from anywhere in the app, plus a standalone chat page and profile embed for visitors. The agent answers grounded in your memos via retrieval-augmented generation, and for the library owner, can take action too: creating, updating, and searching memos as part of the conversation.
Bring your own AI provider key or use MythOS's platform credits; either way, the same agent, the same context, the same conversation.
Features
- 📝Chat — a personal AI agent that reads, writes, and researches inside your library
- 📝BYOK — connect your own AI provider key, or use MythOS platform credits
- 📝Conversation Starters — creator-configured prompts that greet a visitor's first message
- 📝Chat Settings — system prompts, model selection, visitor access, budgets, and rate limits
- 📝Premade Chats — shareable links that open your chat mid-conversation with a preset kickoff and memo scope
Getting Started
Open the sparkles button from anywhere in MythOS to start chatting, or configure your agent at Settings → AI and Settings → Chat.