Datagusto SDK for LangChain / LangGraph / DeepAgents

datagusto-sdk connects agents built with LangChain's create_agent or DeepAgents' create_deep_agent to Datagusto.

Requirements

• Python 3.10 or later
• langchain 1.3+, langchain-core 1.4+, langgraph 1.2+
• deepagents 0.3+ if you use create_deep_agent
• A Datagusto server URL and an agent API key

Installation

pip install datagusto-sdk

Usage

Add DatagustoSafetyMiddleware to your agent's middleware.

With create_agent

from langchain.agents import create_agent
from langgraph.checkpoint.memory import InMemorySaver
from datagusto_langgraph import DatagustoSafetyMiddleware

agent = create_agent(
    model=model,
    tools=tools,
    middleware=[
        DatagustoSafetyMiddleware(
            server_url="https://your-datagusto-server.example.com",
            api_key="your-api-key",
        )
    ],
    checkpointer=InMemorySaver(),
)

agent.invoke(
    {"messages": [{"role": "user", "content": "Summarize today's tickets."}]},
    config={"configurable": {"thread_id": "thread-1"}},
)

With create_deep_agent

from deepagents import create_deep_agent
from datagusto_langgraph import DatagustoSafetyMiddleware

agent = create_deep_agent(
    tools=tools,
    system_prompt="You are a careful research assistant.",
    middleware=[
        DatagustoSafetyMiddleware(
            server_url="https://your-datagusto-server.example.com",
            api_key="your-api-key",
        )
    ],
)

Use the same thread_id for every turn of a conversation.

Configuration

Instead of passing server_url and api_key, you can set environment variables:

export DATAGUSTO_SAFETY_SERVER_URL="https://your-datagusto-server.example.com"
export DATAGUSTO_SAFETY_API_KEY="your-api-key"

DatagustoSafetyMiddleware()

License

MIT — see the LICENSE file.