@arelis-ai/governance-cli

Command-line tool for bootstrapping, provisioning, connecting, validating, and diagnosing AI infrastructure and services used by the Arelis SDK runtime.

Installation

npm install -g @arelis-ai/governance-cli
# or
pnpm add -g @arelis-ai/governance-cli
# or
yarn global add @arelis-ai/governance-cli

Quick Start

# Initialize a new Arelis project
arelis init

# Connect to control plane
arelis connect

# Validate configuration
arelis validate

# Run diagnostics
arelis doctor

Commands

Core Commands

| Command | Description | | ------------------- | ------------------------------------------- | | arelis init | Initialize a new Arelis project | | arelis connect | Connect to Arelis control plane | | arelis provision | Provision infrastructure via IaC | | arelis validate | Validate configuration and connectivity | | arelis doctor | Run diagnostics and get remediation hints | | arelis config | View and manage configuration | | arelis status | Summarize config, auth, and endpoint health | | arelis destroy | Destroy provisioned infrastructure | | arelis completion | Generate shell completions |

Auth Commands

| Command | Description | | -------------------- | ------------------------------------------- | | arelis auth login | Store control plane credentials for CLI use | | arelis auth logout | Remove stored credentials | | arelis auth status | Show auth status | | arelis auth whoami | Show local CLI identity context |

Domain Commands

| Command | Description | | ----------------------- | ---------------------------------- | | arelis mcp add | Add an MCP server to configuration | | arelis mcp list | List configured MCP servers | | arelis mcp test | Test MCP server connectivity | | arelis knowledge add | Add a knowledge base | | arelis knowledge list | List configured knowledge bases | | arelis knowledge test | Test knowledge base connectivity | | arelis models test | Test model provider endpoint | | arelis agents run | Run a single-step agent prompt |

Config Commands

| Command | Description | | ---------------------------------- | ----------------------------- | | arelis config get <path> | Read a configuration value | | arelis config set <path> <value> | Update a configuration value | | arelis config patch <json> | Patch configuration with JSON |

Provider Commands

| Command | Description | | --------------------------- | ------------------------------------- | | arelis hf auth login | Configure HuggingFace authentication | | arelis hf endpoint create | Create HuggingFace Inference Endpoint | | arelis aws auth status | Check AWS credentials | | arelis aws provision | Provision AWS infrastructure | | arelis gc auth login | Configure Google Cloud authentication | | arelis gc provision | Provision GCP infrastructure | | arelis azure auth login | Configure Azure authentication | | arelis azure provision | Provision Azure infrastructure |

Global Flags

| Flag | Description | | ------------------- | ---------------------------------------------- | | --config <path> | Override config file location | | --json | Output in JSON format | | --output <format> | Output format (text, json, yaml, table, sarif) | | --verbose | Detailed output | | --no-color | Plain output without colors | | --yes | Auto-confirm prompts | | --dry-run | Print actions without executing | | --profile <name> | Select configuration profile |

Configuration

Arelis CLI uses arelis.config.ts (or .json) for project configuration. It also supports profile-specific overrides and layered configs:

• Project base: arelis.config.ts / arelis.config.json
• Project profile: arelis.config.<profile>.ts / arelis.config.<profile>.json
• Local overrides: .arelis/config.json (and .arelis/config.<profile>.json)
• Global overrides: ~/.config/arelis/config.json (or %APPDATA%/arelis/config.json on Windows)

import type { ArelisProjectConfig } from '@arelis-ai/governance-cli';

const config: ArelisProjectConfig = {
  project: {
    name: 'my-ai-app',
    environment: 'prod',
  },
  controlPlane: {
    endpoint: 'https://aims.company.com',
    apiKeyRef: 'env:ARELIS_API_KEY',
  },
  audit: {
    sink: 'http',
    endpoint: 'https://audit.company.com/ingest',
    mode: 'hash',
  },
  telemetry: {
    otelEndpoint: 'http://otel-collector:4318',
    serviceName: 'my-ai-app',
  },
  models: {
    defaultRoute: 'primary',
    providers: {
      huggingface: [
        {
          id: 'support-llm',
          kind: 'hf-inference-endpoint',
          endpoint: 'https://xxx.endpoints.huggingface.cloud',
          authRef: 'env:HUGGINGFACE_TOKEN',
          governance: {
            dataResidency: 'EU',
            lifecycle: 'approved',
          },
        },
      ],
    },
  },
};

export default config;

Plugins

Load plugins by listing them in your config or via ARELIS_PLUGINS:

{
  "cli": {
    "plugins": ["@arelis-ai/governance-cli-plugin-example"]
  }
}

Plugins may register additional commands and validation checks by exporting a default plugin object.

Secret References

Arelis CLI never stores secrets directly. Instead, use references:

| Format | Example | Description | | ---------------- | ----------------------------- | -------------------- | | env:VAR | env:ARELIS_API_KEY | Environment variable | | file:/path | file:/etc/secrets/token | File path | | vault:path#key | vault:secret/arelis#api_key | HashiCorp Vault (v1) |

Infrastructure as Code

The CLI supports multiple IaC tools:

Terraform

arelis provision --iac terraform --stack audit-pipeline --cloud aws

CDK

arelis aws provision audit-pipeline --iac cdk

Helm

arelis provision --iac helm --stack otel-collector

Provider Integration

HuggingFace

# Login
arelis hf auth login

# Create endpoint
arelis hf endpoint create \
  --name support-llm \
  --repo my-org/model \
  --region eu \
  --instance nvidia-a10g

# Check status
arelis hf endpoint status support-llm

AWS

# Check credentials
arelis aws auth status

# Provision infrastructure
arelis aws provision audit-pipeline --iac cdk --region us-east-1

# Deploy model
arelis aws deploy sagemaker --model s3://bucket/model.tar.gz

# Register endpoint
arelis aws register --id my-model --endpoint https://...

Google Cloud

# Login
arelis gc auth login

# Check status
arelis gc auth status

# Provision infrastructure
arelis gc provision audit-pipeline --iac terraform --project my-project

# Deploy model
arelis gc deploy vertex --name my-model

Azure

# Login
arelis azure auth login

# Check status
arelis azure auth status

# Provision infrastructure
arelis azure provision audit-pipeline --iac bicep

# Deploy model
arelis azure deploy azure-ml --name my-model

CI/CD Integration

# .github/workflows/validate.yml
name: Validate Arelis Config
on: [push, pull_request]
jobs:
  validate:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - run: npm install
      - run: npx arelis validate --json
        env:
          ARELIS_API_KEY: ${{ secrets.ARELIS_API_KEY }}

Exit Codes

| Code | Description | | ---- | ---------------------------------------------- | | 0 | Success | | 1 | Error (validation failure, connectivity issue) |

Development

# Build
pnpm build

# Test
pnpm test

# Lint
pnpm lint

# Type check
pnpm typecheck

License

MIT