raga-canvas-cli 0.1.5

CLI tool for interacting with Raga Canvas no-code agent deployment platform

Raga Canvas CLI

A command-line interface for interacting with the Raga Canvas no-code agent deployment platform.

Installation

pip install raga-canvas-cli

Quick Start

1. Initialize a new Canvas project

canvas init

This creates the folder structure, sample schemas, and starter configuration files.

2. Login to your Canvas platform

canvas login

This will prompt for your API base URL and authentication token, storing them securely.

3. List your projects

canvas list projects

4. List agents in a project

canvas list agents --project=<project-id>

5. Deploy an agent

canvas push <agent-name> <project-id>

Commands

canvas init

Creates the folder structure, sample schemas, and a starter canvas.yaml, environments/dev.yaml. Adds .canvasrc to .gitignore.

canvas login

Stores a profile (API base + token) in ~/.canvasrc (OS keychain if available).

canvas list projects

Lists projects from the platform API.

canvas list agents --project={projectId}

Lists agents for a specific project.

canvas push <agent> <projectid>

Pushes the effective agent definition to the target project via platform API:

• Uploads bundle to registry (if configured)
• Registers the version
• Applies metadata (owners/tags)
• Does not set or read secret values—only validates presence by name

Directory Structure

When you run canvas init or pull from a project, the following structure is created:

my-canvas-repo/
├── canvas.yaml                 # workspace config
├── .canvasrc                   # local profiles (ignored by VCS)
├── agents/
│   ├── test-agent/
│   │   ├── agent.yaml          # core definition (env-agnostic)
│   │   ├── README.md
│   │   ├── prompts/
│   │   │   ├── system.txt
│   │   │   └── fewshot.md
│   │   ├── evals/
│   │   │   └── smoke.yaml
│   │   └── agent.lock.json     # generated by `canvas lock/package`
│   └── claim-router/
│       └── agent.yaml
├── tools/
│   └── fetch-rules.yaml
├── datasources/
│   └── rules-db.yaml
└── environments/
    ├── dev.yaml
    ├── stage.yaml
    ├── prod.yaml
    └── customer-acme.yaml

Configuration

The CLI uses the following configuration files:

• ~/.canvasrc - Global user configuration and profiles
• canvas.yaml - Workspace configuration
• environments/*.yaml - Environment-specific overrides

Development

To set up for development:

git clone https://github.com/raga-ai/raga-canvas-cli
cd raga-canvas-cli
pip install -e ".[dev]"

Run tests:

pytest

License

MIT License - see LICENSE file for details.