langgraph-react-agent

LangGraph ReAct Agent

A ReAct agent using LangGraph's prebuilt create_react_agent with math tools, running on Flyte.

What it does

• Creates a ReAct (Reason + Act) agent with OpenAI and LangGraph
• Defines simple math tools (add, multiply) with @flyte.trace for observability
• The agent reasons about which tool to call, observes results, and loops until it has an answer

Setup

cd tutorials/starter-examples/langgraph-react-agent

uv venv .venv --python 3.11
source .venv/bin/activate

uv pip install -r requirements.txt

Flyte Cluster (for remote runs)

To run remotely, configure your Flyte cluster endpoint:

flyte create config \
    --endpoint <your-endpoint> \
    --auth-type headless \
    --builder remote \
    --domain development \
    --project flytesnacks

Don't have a cluster? Request access at flyte.org.

Run

Remote:

uv run flyte run langgraph_react_agent.py agent --request "What is 12 * 7 plus 3?"

Local:

uv run flyte run --local langgraph_react_agent.py agent --request "What is 12 * 7 plus 3?"

Local with TUI:

The TUI gives you a live interactive terminal dashboard showing each task's status, logs, and outputs as the run progresses.

uv pip install textual

uv run flyte run --local --tui langgraph_react_agent.py agent --request "What is 12 * 7 plus 3?"

Browse past local runs in TUI:

uv run flyte start tui

Fetch remote run output

After a remote run completes, you can fetch the output with flyte.remote:

import flyte
from flyte.remote import Run

flyte.init_from_config()

runs = list(Run.listall(task_name="langgraph_env.agent", sort_by=("created_at", "desc"), limit=1))
run = Run.get(runs[0].name)
print(f"Phase: {run.phase}")
if run.phase.name == "SUCCEEDED":
    print(run.outputs())

Requirements

• OPENAI_API_KEY secret configured on your Flyte cluster