Getting Started
Prepare a compatible environment
First prepare the infrastructure, then validate the business functions. Source requirements: Python 3.11+, uv, and the corresponding directories EasyRemote/, EasyNet-Cli/, EasyNet-Axon/ alongside. The gallery uv configuration directly references adjacent SDK source code, with dependency range easynet-sdk>=0.142.22,<0.143. Requires an existing compatible EasyNet Runtime and native SDK dependencies. The host uses a Unix socket; node.serve() only starts the Python-provided process and deploys the function.
First complete login, device Join, and start using the Runtime Quick Start steps. Then verify local status with easynet runtime status. Cross-device invocation also requires the target to be visible and authorized to call. The /path/to/EasyRemote below is the source clone directory to be replaced.
Run through a sample pricing rule
In the pricing case directory:
cd /path/to/EasyRemote/gallery/projects/00_basic_remote_math
uv sync
uv run easyremote doctor
uv run python node.pyThe host will print runtime connection, Local active, and publication-pending status. Keep this terminal open; in a second terminal, navigate to the same directory:
uv run python client.pyWith 48 team seats and annual billing, subtotal_cents=2246400, discount_cents=336960, total_cents=1909440 should be observed. This validates local registration and call chain; then replace with your own calculation, keeping the signature and adding business-scope checks.
Install a network library that contains only interfaces
To experience the interface package, in the network library example directory run:
cd ../08_network_native_python_library
uv sync
export EASYREMOTE_LIBRARY_ROOT="$PWD/.libraries"
uv run easyremote add library.json
uv run python inspect_library.pyYou should see easyremote.silan.lotus, two signatures, and four interface files. Then in separate terminals, run node.py and client.py with the same library root directory on both ends; filter to leave two invoice-related records and output lowercase mappings. The current algorithm is token-item matching with fixed-text transformation. This interface can continue to be used when replacing with a real model.
No comments yet

