capsule

capsule is a small tool with a simple but noble goal; to make deployment of CosmWasm contracts easier.

Firstly we are targeting Terra as the Sponsor User chain as we build out the capsule tool. Eventually we aim to make capsule one of the tools of choice for deploying CosmWasm contracts on all chains!

Installation

Install from pypi

pip install -i https://test.pypi.org/simple/ capsule

We have published the package also under capsule_terra. This is temporary until a final package namespace is chosen.

pip install -i https://test.pypi.org/simple/ capsule_terra

pip install -i https://test.pypi.org/simple/ capsule_terra

Install locally

Git clone the project and change into its parent directory.

git clone
cd 

Install the project using setup tools.

python setup.py install

Access the tool using the command line.

capsule -h

Note: If you have an issue with the above an the command line tool, depending on your platform you will need to prepare a standard distribution and install that. To do so here is two commands together:

python setup.py sdist && pip install dist/capsule-0.0.0.tar.gz

Available Commands and Usage

Deploy - Deploy a given cosm wasm contract artifact to a chain of your choice

capsule deploy -h
usage: 
    $ capsule deploy -p ./my_contract.wasm -c columbus-5
    $ capsule deploy --path ./artifacts/my_contract.wasm --chain tequila-0004
    $ capsule deploy --path ./artifacts/my_contract.wasm --initmsg {'count':3}

Helper tool which enables you to programatically deploy a Wasm contract artifact to a chain as a code object and instantiate it

Configuration

The capsule tool offers the ability to store details you need in a configuration file using the toml format.

The config file by default is located in a capsule specific hidden directory at the home dir level: ~/.capsule It is possible also to specify the path to a custom config file using the CAPSULE_CONFIG_FILE environment variable.

Something to be explored is also enabling both the Mnemonic and the chain to deploy too as env vars also. If this was to happen the order of priority would then become Credentials in the environment -> Config file in the environment -> Default or specified config file. Following this pattern in theory should make this tool very easy to use in CI/CD as a given user can just specify the Mnemonic and chain ID for as secrets in the job for a quick start.

CI/CD

This project uses Github Actions to perform automatic testing on each push and PR as well as a deployment to both test and prod pypi.

Some notes: When deploying to pypi, a version number can only be deployed once! All subsequent deployments will get a 400 and the job will fail.

As a procedure we should update the version number when a new deployed build is needed. Or have its patch versions done automatically.

Lastly the production build which pushes to Pypi only works with a tagged commit.