Skip to main content
This shared setup guide is for the public starter projects referenced by the Practitioner, Builder, and Contributor paths.

​
What this setup is for

  • reading Prompthon Agentic Lab locally
  • running the current repo-owned starter code
  • verifying example projects before or after contribution work
The current sample projects are intentionally lightweight Python starters. They are small enough to inspect quickly and validate without a framework-specific toolchain.

​
Prerequisites

  • Git
  • Python 3.9 or newer
  • a terminal on macOS, Linux, or WSL

​
Clone the repository

git clone https://github.com/Prompthon-IO/agent-systems-handbook.git
cd agent-systems-handbook

​
Optional virtual environment

The current starter checks use only the Python standard library, but an isolated environment is still a good default: 
python3 -m venv .venv
source .venv/bin/activate
python3 --version

​
Verify the current starter code

Run the repo-level smoke check: 
python3 scripts/verify_example_projects.py
This verifies the current starter code under:
  • patterns/examples/
  • systems/examples/
  • ecosystem/examples/
  • case-studies/examples/

​
What to expect

  • The check validates the current code sketches and example flows.
  • The current projects are still documented as starter examples, not full production applications.
  • Some starters demonstrate state shape or control flow without shipping the surrounding transport, persistence, or framework runtime yet.

​
If you are contributing

  • Keep example projects in lane-local examples/ folders.
  • Declare project status as starter, partial, or runnable.
  • If you change executable starter code, update scripts/verify_example_projects.py in the same change.