Report Abuse

Basic Information

ZerePy is an open-source Python framework for building and deploying autonomous agents that interact with social platforms and blockchains. It is a modularized extraction of the Zerebro backend designed to let developers launch agents with similar core capabilities, driven by multiple large language models. The repository provides a CLI-driven platform for configuring connections, loading agents defined as JSON files, and running agent loops or single actions. It targets integrations with social networks such as X (Twitter), Farcaster, Echochambers and Discord, and supports on-chain activity on Solana and EVM chains through GOAT and Monad. The project documents requirements such as Python 3.10+ and Poetry and includes a Replit template for quick testing. Developers can fine-tune models for creative outputs and store sensitive keys in environment files for runtime operation.

Links

App Details

Features
ZerePy exposes a command-line interface for managing agents and connections, a modular connection system for plugging in LLMs and external services, and built-in blockchain integration. Language model adapters include OpenAI, Anthropic, EternalAI, Ollama, Hyperbolic, Galadriel and XAI. Social integrations support posting, reading and replying on X, posting and replying on Farcaster, interacting with Echochambers rooms, and common Discord channel operations. Onchain capabilities cover Solana actions and EVM interactions via GOAT, with plugin support for ERC20, Coingecko, DEX analytics and more. Agents are configured with JSON files that declare bios, traits, examples, connection configs, tasks and scheduling options. The README documents commands such as configure-connection, list-connections, load-agent, start, agent-loop and chat, plus guidance for GOAT plugin installation.
Use Cases
ZerePy helps developers create autonomous social and blockchain-aware agents without building infrastructure from scratch. It centralizes configuration and runtime control via a CLI and structured agent JSON files, enabling repeatable agent behavior such as posting, replying, liking, and timeline reads across multiple social platforms. The GOAT plugin system lets agents access on-chain functionality like token transfers, balance checks, swaps and market data when paired with RPC endpoints and wallet keys. Multiple LLM backends can be selected per agent for flexible model choice or fine-tuning workflows. The project supports rapid prototyping via a Replit template and documents installation with Poetry, easing local development and testing. Sensitive credentials are expected to live in environment files and plugins can be added to extend functionality.

Please fill the required fields*