Skip to main content

​
AI Agents API Guide

Learn how to work with AI agents using the Wacht Rust SDK.

​
Prerequisites

Before using any API methods, you must initialize the SDK: 
use wacht::init_from_env;

#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
    // Initialize from environment variables
    // Requires: WACHT_API_KEY and WACHT_FRONTEND_HOST
    init_from_env().await?;

    // Now you can use the API
    Ok(())
}

​
List AI Agents

Get all AI agents in your deployment.

​
Basic List

use wacht::api::ai;

// Fetch all agents (default pagination)
let agents = ai::list_agents()
    .send()
    .await?;

for agent in agents.data {
    println!("{}: {}", agent.name, agent.description.unwrap_or_default());
}

​
With Filters

use wacht::api::ai;

// Fetch active agents with pagination
let agents = ai::list_agents()
    .limit(50)
    .offset(0)
    .is_active(true)
    .search("support")
    .send()
    .await?;

for agent in agents.data {
    println!("{} (active: {})", agent.name, agent.is_active);
}

​
Get AI Agent

Retrieve a specific AI agent by ID. 
use wacht::api::ai;

let agent = ai::fetch_agent("agent_123")
    .send()
    .await?;

println!("Agent: {}", agent.name);
println!("Description: {:?}", agent.description);

​
Get Agent Details

Retrieve complete details about an AI agent. 
use wacht::api::ai;

let details = ai::fetch_agent_details("agent_123")
    .send()
    .await?;

println!("Name: {}", details.name);
println!("Configuration: {:?}", details.configuration);
println!("Tools: {:?}", details.tool_ids);
println!("Knowledge Bases: {:?}", details.knowledge_base_ids);

​
Create AI Agent

Create a new AI agent with custom configuration.

​
Basic Agent Creation

use wacht::api::ai;
use wacht::models::CreateAiAgentRequest;

let agent = ai::create_agent(CreateAiAgentRequest::new(
    "Support Bot".to_string(),
    "gpt-4".to_string()
))
.send()
.await?;

println!("Created agent: {}", agent.id);

​
Agent with Configuration

use wacht::api::ai;
use wacht::models::CreateAiAgentRequest;

let agent = ai::create_agent(CreateAiAgentRequest {
    name: "Support Bot".to_string(),
    model: "gpt-4".to_string(),
    description: Some("Customer support assistant".to_string()),
    system_prompt: Some("You are a helpful support assistant.".to_string()),
    temperature: Some(0.7),
    max_tokens: Some(2000),
    tools: Some(vec!["tool_123".to_string()]),
    knowledge_bases: Some(vec!["kb_789".to_string()]),
})
.send()
.await?;

​
Update AI Agent

Modify an existing AI agent’s configuration. 
use wacht::api::ai;
use wacht::models::UpdateAiAgentRequest;

let updated = ai::update_agent(
    "agent_123",
    UpdateAiAgentRequest {
        name: Some("Updated Support Bot".to_string()),
        description: Some("Enhanced customer support".to_string()),
        model: Some("gpt-4-turbo".to_string()),
        system_prompt: Some("Updated instructions.".to_string()),
        temperature: Some(0.6),
        is_active: Some(true),
        ..Default::default()
    }
)
.send()
.await?;

​
Delete AI Agent

Permanently delete an AI agent. 
use wacht::api::ai;

ai::delete_agent("agent_123")
    .send()
    .await?;

​
Builder Methods

​
ListAgentsBuilder

  • limit(i32) - Number of results to return (max 100)
  • offset(i32) - Number of results to skip
  • is_active(bool) - Filter by active status
  • search(&str) - Search agents by name or description

​
Request Models

​
CreateAiAgentRequest

  • name (String) - Agent name (required)
  • model (String) - Model to use (required)
  • description (Option<String>) - Agent description
  • system_prompt (Option<String>) - System instructions
  • temperature (Option<f64>) - Temperature (0.0 - 2.0)
  • max_tokens (Option<i32>) - Maximum tokens in response
  • tools (Option<Vec<String>>) - Associated tool IDs
  • knowledge_bases (Option<Vec<String>>) - Associated knowledge base IDs

​
UpdateAiAgentRequest

All fields are optional - only include what you want to change:
  • name (Option<String>)
  • description (Option<String>)
  • model (Option<String>)
  • system_prompt (Option<String>)
  • temperature (Option<f64>)
  • max_tokens (Option<i32>)
  • tools (Option<Vec<String>>)
  • knowledge_bases (Option<Vec<String>>)
  • is_active (Option<bool>)

​
Agent Configuration

Agents are configured using direct fields rather than a nested configuration object: 
CreateAiAgentRequest {
    name: "Support Bot".to_string(),
    model: "gpt-4".to_string(),
    system_prompt: Some("You are a helpful assistant.".to_string()),
    temperature: Some(0.7),
    max_tokens: Some(2000),
    tools: Some(vec!["tool_123".to_string()]),
    knowledge_bases: Some(vec!["kb_789".to_string()]),
}

​
Error Handling

All SDK methods return a Result<T, Error>: 
use wacht::{Error, api::ai};

match ai::fetch_agent("invalid_id").send().await {
    Ok(agent) => println!("Agent: {}", agent.name),
    Err(Error::Api { status, message, .. }) => {
        eprintln!("API Error {}: {}", status, message);
    }
    Err(Error::Request(e)) => {
        eprintln!("Network error: {}", e);
    }
    Err(e) => {
        eprintln!("Other error: {}", e);
    }
}