langchain

Predactiv MCP — LangChain AI Agent

This sample shows how to connect an AI agent to the Predactiv MCP server using LangChain and langchain-mcp-adapters.

The agent authenticates to Predactiv, discovers the available tools from the MCP server at runtime, and lets a large language model call those tools to work with your Predactiv data platform (audiences, datasets, datasources, destinations, and more).

How it works

1. Authenticate — oauth_token_lib.auth_headers() exchanges your CLIENT_ID and CLIENT_SECRET for an OAuth2 bearer token (client-credentials flow) and returns the Authorization header.
2. Connect & discover tools — MultiServerMCPClient connects to the Predactiv MCP server over HTTP and fetches its tools dynamically. No tools are hardcoded, so your agent always sees the latest set the server exposes.
3. Run the agent — the MCP tools are bound to an LLM via create_agent, and the agent is invoked with a natural-language goal.

Prerequisites

• Python 3.9+
• A Predactiv account with API credentials (CLIENT_ID and CLIENT_SECRET). Contact Predactiv if you don't have these.
• An OpenAI API key — this sample uses gpt-5.4-mini as the agent's LLM. Set it as the OPENAI_API_KEY environment variable.

OpenAI is just the demonstration LLM — not a requirement. This sample is model-agnostic: LangChain works with any chat model that supports tool calling. To use a different LLM, swap the ChatOpenAI(...) client for another LangChain chat model (e.g. ChatAnthropic), install its integration package, set that provider's API key, and update the OPENAI_MODEL constant at the top of agent.py. The Predactiv MCP side is unchanged.

Setup

From this langchain/ directory:

# (recommended) create and activate a virtual environment
python -m venv .venv
source .venv/bin/activate        # on Windows: .venv\Scripts\activate

# install dependencies (also installs the shared oauth-token-lib package)
pip install -r requirements.txt

Configuration

Set the following environment variables:

export CLIENT_ID=your-predactiv-client-id
export CLIENT_SECRET=your-predactiv-client-secret
export OPENAI_API_KEY=your-openai-api-key

Run

python agent.py

The sample's default goal is:

Get list of my audiences from the Predactiv MCP server and return it to me.

Edit the GOAL constant in agent.py to ask the agent to do something else with your Predactiv data.

Troubleshooting

• CLIENT_ID and CLIENT_SECRET environment variables must be set. — Export both variables (see Configuration) before running.
• 401 Unauthorized from the MCP server — Your credentials are invalid or expired. Verify them, or run python ../oauth-token-lib/oauth_token_lib.py to confirm a token can be obtained.
• OpenAI authentication errors — Make sure OPENAI_API_KEY is set and valid.

Learn more

• Predactiv MCP server: https://mcp.predactiv.com
• Model Context Protocol: https://modelcontextprotocol.io