Deep Agents Template

Deployment template for a deep agent built with create_deep_agent(...).

What this template gives you

• A deployable deep agent graph at src/deep_agent/graph.py.
• Explicit workflow prompt (plan, delegate, critique, finalize).
• Two predefined sub-agents (researcher, critic).
• Human-in-the-loop interrupts on execute and write_file.
• A uv-managed local workflow with a small Makefile wrapper and starter tests.

Prerequisites

• An API key for your model provider (Anthropic by default)
• A LangSmith account (Plus plan or higher) to deploy

Quickstart

1. Sync the project and configure environment:

uv sync
cp .env.example .env

2. Start the dev server:

uv run langgraph dev

3. Deploy to LangSmith:

uv run langgraph deploy

See the CLI docs for deploy options.

To set up CI instead, push this repo to GitHub and configure your deployment through the LangSmith UI.

Tests and lint

make test
make integration-tests
make lint
make format

Integration tests are skipped unless ANTHROPIC_API_KEY is set.

Reference docs

• Deep Agents overview: https://docs.langchain.com/oss/python/deepagents/overview
• Deep Agents quickstart: https://docs.langchain.com/oss/python/deepagents/quickstart
• LangSmith CLI: https://docs.langchain.com/langsmith/cli