agent-tree

🌳 agent-tree

The missing visualizer for AI Agents. Turn messy terminal logs into beautiful, interactive trees.

agent-tree is a zero-config, terminal-first observability tool. It captures the complex "reasoning loop" of AI agents and renders them as clean, hierarchical trees in your console. No API keys, no web dashboards, no fluff.

Why agent-tree?

📸 Why use agent-tree?

🚀 Quick Start

1. Automatic (LangChain/LangGraph)

Just pass the TraceCallback into your agent's config. It automatically catches the end of the run and draws the tree.

from agent_tree import TraceCallback

# Your agent logic...
agent.invoke(
    {"messages": [("user", "Is 50 shares of Apple under $15,000?")]},
    config={"callbacks": [TraceCallback(session_name="Budget Check")]}
)

2. Manual (Post-Execution)

If you already have a result object, you can visualize it instantly.

from agent_tree import AgentTrace

at = AgentTrace(result_object)
at.show()

✨ Key Features

• Instant Visuals: Built on top of the Rich library for high-performance terminal UI.
• Root-Only Tracing: Smart logic prevents "log spam" by only showing the final, complete execution tree.
• Framework Agnostic: Works perfectly with LangChain but can be adapted for any JSON-like agent output.

🛠 Installation

pip install agent-tree

Github Repository

You can find the full source code and contribute on GitHub.