Metadata-Version: 2.4
Name: canyonos
Version: 0.2.2rc1
Summary: Open-source platform for multi-agent apps: per-agent scaling, tracing and cost, in plain Python
License-Expression: AGPL-3.0-only
Project-URL: Homepage, https://canyonos.ai
Project-URL: Documentation, https://canyonos.readthedocs.io/en/stable/
Project-URL: Source, https://github.com/CanyonCodeCoreAI/canyonos
Project-URL: Changelog, https://canyonos.ai/changelog
Keywords: multi-agent,ai-agents,llm,observability,opentelemetry,llm-cost
Classifier: Intended Audience :: Developers
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
Classifier: Topic :: System :: Monitoring
Classifier: Programming Language :: Python :: 3
Requires-Python: >=3.10
Description-Content-Type: text/markdown
Requires-Dist: boto3[crt]>=1.43.109
Requires-Dist: pyfiglet
Requires-Dist: python-dotenv
Requires-Dist: pyyaml
Requires-Dist: rich
Requires-Dist: ruamel.yaml

# canyonos CLI

The `canyonos` command line for [CanyonOS](https://canyonos.ai): port a project, deploy it, and manage the running deploy.

## Install

```bash
uv tool install canyonos
canyonos doctor   # check Docker and Compose are available
```

## Use

Converting an existing workflow? Run `canyonos convert` in its folder first; your coding
agent writes a CanyonOS version of it into a `.canyon/` folder.

```bash
cd your-project
canyonos deploy   # build and launch the workflow, then open the dashboard
canyonos stop     # stop the deploy and remove its containers and files
```

Run `canyonos -h` for every command.

## Develop

Run the CLI from this checkout with `uv run canyonos <command>`. Setup is in
[docs/contributing/DEVELOPMENT.md](https://github.com/CanyonCodeCoreAI/canyonos/blob/main/docs/contributing/DEVELOPMENT.md), and how the CLI
works is in [docs/architecture/CLI.md](https://github.com/CanyonCodeCoreAI/canyonos/blob/main/docs/architecture/CLI.md).
