branchy-chat 0.1.2

Branching conversation trees for AI chat. Store, retrieve, and route conversations with tree-based memory.

branchy-chat

Store conversation messages as a tree structure, retrieve relevant context traces, and let your LLM automatically route new prompts to the right branch.

• 🌳 Tree-based conversation memory
• 🔍 Retrieve relevant conversation traces
• 🧠 LLM-powered branch routing
• 📦 Publishable pip package

Installation

pip install branchy-chat

Usage

from branchy import BranchingChat

chat = BranchingChat(
	model="openai/gpt-4o-mini",
	api_key="<your-api-key>",
	api_base="http://localhost:4000",  # Optional, useful for LiteLLM proxy
)

message_id_1 = chat.add_message("How do I build an index in Postgres?")
message_id_2 = chat.add_message("What if I need it concurrently?")

# Display the tree structure
print(chat.show_tree())

# Display a subtree starting from a specific message
print(chat.show_subtree(message_id_1))

LiteLLM compatibility

BranchingChat supports LiteLLM-style parameters:

• model
• api_key
• api_base

This means you can point it at a LiteLLM proxy endpoint by setting api_base.

API

• BranchingChat(*, model, api_key=None, api_base=None, **lm_kwargs) : Initialize using LiteLLM-compatible arguments.
• BranchingChat.add_message(user_message: str) -> str : Routes a message to the best branch, stores it in the tree, and returns the new message_id.
• BranchingChat.show_tree() -> str : Renders user messages in an ASCII tree format and returns the formatted string.
• BranchingChat.show_subtree(message_id: str) -> str : Renders a subtree starting from a given message_id and all its child branches.