This repository serves as a community-driven hub for LangChain artifacts, offering a centralized resource for sharing and discovering prompts, chains, and agents to build complex LLM applications. It aims to be a community-driven alternative to Hugging Face Hub for LangChain-specific components.

How It Works

LangChainHub organizes reusable components like prompts, chains, and agents into a structured directory. Users can load these artifacts directly into their LangChain applications using specific load_prompt, load_chain, or initialize_agent functions with lc:// URIs. This approach simplifies the integration of pre-built, community-vetted components, promoting code reuse and accelerating LLM application development.

Quick Start & Requirements

• To load artifacts: from langchain.prompts import load_prompt; prompt = load_prompt('lc://prompts/path/to/file.json') (and similar for chains/agents).
• Requires LangChain library.
• Contribution via Pull Requests or GitHub Issues.

Highlighted Details

• Collection of prompts, with plans to expand to chains and agents.
• Artifacts are organized by use case within respective directories.
• Each artifact sub-directory includes a README for usage guidance.

Maintenance & Community

• This repo is being replaced by the hosted LangChain Hub Product at https://smith.langchain.com/hub.

Licensing & Compatibility

• License not specified in the README.

Limitations & Caveats

The repository is being deprecated in favor of a hosted product, indicating potential future maintenance issues or a shift in focus away from this specific GitHub repository. Agent Executors are listed as "Coming soon."