Using private python packages in robots
Recipe for accessing private PyPI packages with credentials in Robocorp Vault
Recipe for using private packages securely with credentials in Robocorp Vault.
In the example requirements-private.txt you'll find how to install Python packages from different sources:
- Private PyPI service.
- Private GitHub repository.
- Local source with a directory path accessible by the robot.
Setup & Run
Go to Robocorp Control Room and create a Vault secret with name:
private_pypiand add the following keys:
pypi_usr: username for private PyPI source
pypi_pwd: password for private PyPI source
github_token: Access token for GitHub source
- Link your VS Code Robocorp extension to Control Room so that it has access to the Vault secrets.
- The robot executions with Workforce Agent and Assistant automatically get the secure Vault access.
- Run the bot with VSCode and select a task demonstrating how the private packages get installed and used.
Private PyPI Library Usage
- Installs and uses a library from a private PyPI.
- Requires a Vault secret with name
Private Repo Or Local Source Library Usage
- Installs and uses a library from either a private repository or a local source.
- Requires a Vault secret with the name
- To choose between the sources, just comment out the lines you don't want in the requirements file.
How it works
- The robot.yaml contains an entry called
preRunScriptswhich instructs rcc to execute a script like private-pip-install.py right before the robot run.
- The script receives as parameters the Vault secret store name and the path to the requirements file to be installed by pip, containing private sources for the dependencies.
- The script injects into the environment variables used by the
pip install, which is run to install the private requirements.
- Once the script finishes, the selected task will run and use the previous library installed, thus demonstrating that the Python environment contains the requested dependencies from private sources.