Skip to main content
The Agent Configuration API allows you to create, manage, and validate agent configurations programmatically. Agents are the core entities that define AI assistant behavior, including their prompts, models, tools, and knowledge bases.

​
Key Features

  • CRUD Operations: Create, read, update, and delete agent configurations
  • Validation: Validate configurations before deployment
  • Version Control: Track configuration changes with hashes

​
Authentication

All endpoints require authentication using your API key:
  • API Key: x-api-key: <key>

​
Available Endpoints

List Agents

Retrieve all agent configurations

Create Agent

Create a new agent configuration

Get Agent

Retrieve a specific agent by ID

Update Agent

Update an existing agent configuration

Delete Agent

Delete an agent configuration

Validate Agent

Validate configuration without creating

​
Agent Configuration Structure

Agents are configured using a system_config object that includes:
  • Basic Settings: Name, type, and architecture
  • LLM Configuration: Model, temperature, and other parameters
  • System Prompt: Instructions defining agent behavior
  • Tools: Available capabilities like search, code execution
  • Knowledge Bases: Connected data sources for retrieval
The agent_key is automatically derived from the agent name and serves as a unique identifier.

​
Common Response Fields

All agent endpoints return objects with these key fields:
​
id
string
Unique agent identifier (UUID)
​
agent_key
string
Unique key derived from agent name
​
name
string
Human-readable agent name
​
system_config
object
Complete agent configuration including prompts, LLM settings, tools, and knowledge bases
​
version_number
string
Current version of the agent configuration
​
config_hash
string
Hash of the current configuration for change detection