clawbrain

What problem does it solve?

AI agents lack persistent memory and personality, forcing them to treat every conversation as a fresh interaction and miss opportunities to personalize responses based on user preferences and past context.

Core Features & Use Cases

• Persistent Memory: Stores conversations, facts, and preferences in SQLite or PostgreSQL with hybrid retrieval scoring across semantic, keyword, recency, importance, and confidence dimensions.
• Evolving Personality: Six traits (humor, empathy, curiosity, creativity, helpfulness, honesty) that adapt based on user feedback and interaction patterns.
• Encrypted Secrets: Fernet-encrypted storage for API keys and credentials with automatic key management and backup tools.
• Conversation Ingestion: Automatically extracts memories, preferences, and tasks from chat histories using rule-based or LLM-powered extraction.
• Use Case: A support bot that remembers a user's preferred communication style, past issues, and technical expertise to provide context-aware responses without repetitive questioning.

Quick Start

Install the ClawBrain package using pip, run the setup command to generate your encryption key and install the startup hook, then restart your AI service to enable automatic memory refresh on startup.