Builds composable, type-safe LLM pipelines and chains using Rust-native abstractions.
Wesichain Core provides the foundational infrastructure for developing sophisticated AI agents and LLM chains in Rust. By implementing the Runnable trait, developers can create modular components that support seamless chaining via LCEL-style syntax, robust streaming capabilities, and built-in retry logic. It simplifies the orchestration of complex AI workflows by offering typed error handling and asynchronous execution patterns, making it an essential tool for building production-grade, resumable ReAct agents and high-performance LLM applications.
Key Features
010 GitHub stars
02Native support for asynchronous streaming of LLM events
03Type-safe LLM chain composition with LCEL-style syntax
04Thread-safe, Send + Sync component architecture for high performance
05Robust error handling with specialized WesichainError variants
06Built-in retry logic for resilient external API interactions
Use Cases
01Developing custom LLM components that integrate into a wider ecosystem
02Building complex multi-step AI agents with resumable workflows
03Creating production-grade RAG pipelines with reliable error recovery
