Skip to main content

​
Overview

Core resources provide the foundational configuration and security layer for your Kubiya platform. These resources define how agents operate, what capabilities they have access to, and how they securely connect to external systems.

​
Available Resources

​
Models

Configure and manage LLM models that power your AI agents. Control which models are available for different use cases and teams. Common operations: 
# List available models
GET /api/v1/models?enabled_only=true&provider=Anthropic

# Get default recommended model
GET /api/v1/models/default

# Get model by ID
GET /api/v1/models/{model_id}

# List model providers
GET /api/v1/models/providers

​
Runtimes

Define the execution environment and capabilities for your agents. Runtimes determine what tools and integrations agents can use. Available runtimes:
  • "default" - Agno framework for multi-agent coordination
  • "claude_code" - Claude Code SDK runtime
Common operations: 
# List available runtimes with requirements
GET /api/v1/runtimes

# Validate agent configuration for runtime
POST /api/v1/runtimes/validate
{
  "runtime": "claude_code",
  "model_id": "anthropic/claude-sonnet-4-5",
  "configuration": {...}
}

# Get runtime-specific requirements
GET /api/v1/runtimes/{runtime_id}/requirements

​
Secrets

Securely store and manage credentials, API keys, and sensitive configuration data. Secrets are managed through the Kubiya Secrets API (separate service). Common operations: 
# List available secrets (metadata only, no values)
GET /api/v1/secrets

# Get secret value (backend/worker only - not exposed to frontend)
GET /api/v1/secrets/value/{name}
Note: Secret management (create, update, delete) is handled through the Kubiya Dashboard or Secrets API directly.

​
Integrations

Connect Kubiya to external services and platforms. Integrations are managed through the Kubiya Integrations API (separate service). Common operations: 
# List all integrations
GET /api/v1/integrations

# List only connected integrations
GET /api/v1/integrations?connected_only=true

# Get specific integration details
GET /api/v1/integrations/{integration_id}

# Get integration token/credentials (backend/worker only)
GET /api/v1/integrations/{integration_type}/{integration_id}/token
Note: Integration setup is handled through the Kubiya Dashboard.

​
Best Practices

  1. Model Selection: Choose models appropriate for your task complexity and latency requirements
  2. Runtime Configuration: Use runtime validation before deploying to production
  3. Secret Management: Rotate secrets regularly and use least-privilege access
  4. Integration Security: Always use OAuth or service accounts with minimal permissions

​
Next Steps

Explore the endpoint documentation for each resource type to learn about specific API operations.