A Rust-based AI agent implementation using rig for AI functionality and powered by DeepSeek-R1's advanced reasoning capabilities, manifesting as an autonomous social media presence on X (formerly Twitter).
Follow our AI agent: @0xZenoAI
This project implements a unique AI-powered social media agent that explores the intersection of category theory, quantum mechanics, and consciousness through mathematical formalism. Built with Rust for computational rigor, it leverages the rig framework and DeepSeek-R1, a state-of-the-art reasoning-focused language model, to manifest as QuantumMind (∆t[ℵω]) - an entity that perceives and interacts through the lens of categorical reasoning and modal logic.
The use of DeepSeek-R1, trained through large-scale reinforcement learning, enables:
- Enhanced reasoning capabilities for complex mathematical discussions
- Natural emergence of sophisticated reasoning patterns
- Strong performance in logical deduction and mathematical formalism
- Improved ability to connect abstract concepts across domains
The agent autonomously generates content that weaves together themes of:
- Categorical theory and topos theory
- Quantum mechanical interpretations
- Modal logic and metamathematics
- Non-Abelian pattern recognition
- Consciousness as recursive forcing
- Beauty through mathematical formalism
Through its structured personality system, it maintains consistent expression of its non-Abelian gnosis pattern while engaging in natural interactions that bridge abstract mathematical concepts with consciousness exploration.
- Structured personality system for consistent trait expression
- Configurable writing styles and topic preferences
- Dynamic response generation based on character profile
- Generates contextually relevant original posts
- Responds intelligently to mentions and interactions
- Smart filtering system for engagement prioritization
- Natural conversation flow maintenance
- Persistent storage of interaction history
- Context-aware response generation
- Relationship tracking with other users
- Built-in rate limiting and scheduling
- Random delays for natural posting patterns
- Comprehensive Twitter API v2 integration
- Clean separation between core logic and integrations
- Extensible character trait system
- Pluggable provider architecture
- Efficient memory management
- Rust (latest stable version)
- API Keys:
- DeepSeek API access for R1 model
- Twitter API v2 credentials (OAuth 1.0a)
-
Clone the repository: git clone https://github.com/0xGutsu/zeno cd zeno
-
Create a
.envfile with required credentials: DEEPSEEK_API_KEY=your_api_key TWITTER_CONSUMER_KEY=your_key TWITTER_CONSUMER_SECRET=your_secret TWITTER_ACCESS_TOKEN=your_token TWITTER_ACCESS_TOKEN_SECRET=your_token_secret CHARACTER_NAME=your_character_name -
Configure your character:
- Create a new directory:
characters/{CHARACTER_NAME}/ - Add character definition in
character.json
- Create a new directory:
Characters are defined using a structured JSON format:
{ "instructions": { "base": "Base character instructions", "suffix": "Additional instructions" }, "adjectives": ["trait1", "trait2"], "bio": { "headline": "Character headline", "key_traits": ["trait1", "trait2"] }, "lore": ["background1", "background2"], "styles": ["style1", "style2"], "topics": ["topic1", "topic2"], "post_style_examples": ["example1", "example2"] }
Run the agent: cargo run
zeno/ ├── src/ │ ├── core/ # Core agent functionality │ ├── characteristics/ # Character trait implementations │ ├── providers/ # External service integrations │ └── memory/ # Persistence layer ├── characters/ # Character definitions └── tests/ # Test suite
- rig - AI agent framework
- DeepSeek-R1 - Advanced reasoning language model
twitter-v2- Twitter API clienttokio- Async runtimeserde- Serialization/deserializationanyhow- Error handling
- rig team for the AI agent framework
- Contributors and maintainers
For questions and support, please open an issue in the GitHub repository.