Deploy Migration
Move from praisonai.deploy to the standalone praisonai-deploy package
praisonai-deploy is a standalone tier-2 package. Existing praisonai.deploy imports keep working through compatibility shims.
# New code
from praisonai_deploy import Deploy
# Still works (compatibility shim)
from praisonai.deploy import DeployQuick Start
- Update your import
python from praisonai_deploy import Deploy - Everything else stays the same
python deploy = Deploy.from_yaml("agents.yaml") result = deploy.deploy()
Import Mapping
| Old path | New path |
|---|---|
from praisonai.deploy import Deploy | from praisonai_deploy import Deploy |
from praisonai.deploy import DeployConfig | from praisonai_deploy import DeployConfig |
praisonai.cli.commands.deploy | forwarded to the new Typer subcommands |
Compatibility Coverage
The extraction preserves existing imports through alias shims:
praisonai.deployforwards topraisonai_deploy.praisonai.cli.commands.deployforwards to the new Typer subcommands.- Legacy scheduler imports continue to resolve.
praisonai-codecalls into deploy through a bridge, without a hard dependency.
These are verified by regression gates:
scripts/check_c14_deploy_imports.shtest_c14_deploy_backward_compat.py(18 passing)praisonai-deployunit tests (115 passing)
Install
pip install praisonai-deploypip install "praisonai[deploy]"The full praisonai umbrella keeps the wrapper shims installed, so existing scripts run unchanged.
Best Practices
Migrate imports gradually
The shims mean you can update imports file by file without breaking anything in between.
Install the standalone package for slim deploys
pip install praisonai-deploy gives you the deploy surface without pulling the full wrapper.