Skip to main content

​
Authentication and Initialization

The SDK uses cookie-based authentication with Bearer token format. The API key is passed as a cookie named access_token with value Bearer {api_key}. 
from celesto.sdk import CelestoSDK

# Using explicit API key
client = CelestoSDK(api_key="sk-your-api-key-here")

​
ToolHub

​
list_tools

Retrieves the catalog of available managed tools. 
tools = client.toolhub.list_tools()
# Returns: {"tools": [{"name": "get_weather", "description": "...", ...}, ...]}

​
run_weather_tool

Executes the weather tool to retrieve current weather data for a specified city. 
weather = client.toolhub.run_weather_tool("London")
# Returns: {"error": None, "data": {...weather data...}}

​
list_google_emails

Lists all emails from the user’s Gmail account. 
emails = client.toolhub.run_list_google_emails(limit=5)

​
Deployment

from pathlib import Path
from celesto.sdk import CelestoSDK

# Initialize SDK
client = CelestoSDK(api_key="sk-your-api-key")

# === ToolHub Operations ===

# List available tools
tools = client.toolhub.list_tools()
print(f"Available tools: {len(tools['tools'])}")

# Execute weather tool
weather = client.toolhub.run_weather_tool("San Francisco")
print(f"Weather: {weather}")

# Execute Gmail tools (requires Google OAuth setup)
emails = client.toolhub.run_list_google_emails(limit=3)
print(f"Recent emails: {len(emails)}")

client.toolhub.run_send_google_email(
    to="[email protected]",
    subject="Hello from Agentor",
    body="This email was sent via CelestoSDK"
)

# === Deployment Operations ===

# Deploy an agent
deployment = client.deployment.deploy(
    folder=Path("./my-weather-agent"),
    name="weather-agent-prod",
    description="Production weather assistant",
    envs={
        "OPENAI_API_KEY": "sk-...",
        "LOG_LEVEL": "INFO"
    }
)

print(f"Deployment ID: {deployment['id']}")
print(f"Status: {deployment['status']}")

# List all deployments
all_deployments = client.deployment.list()
for dep in all_deployments:
    print(f"{dep['name']}: {dep['status']} (created: {dep['created_at']})")

​
deploy

The Deployment client manages the agent deployment lifecycle, including packaging, uploading, and listing deployed applications. Signature: deploy(folder: Path, name: str, description: Optional[str] = None, envs: Optional[dict[str, str]] = None) -> dict Parameters:
  • folder (Path): Path to directory containing agent code
  • name (str): Deployment name (must be unique)
  • description (str, optional): Human-readable deployment description
  • envs (dict[str, str], optional): Environment variables as key-value pairs
from pathlib import Path

result = client.deployment.deploy(
    folder=Path("./my-agent"),
    name="weather-agent-v1",
    description="Weather assistant",
    envs={"OPENAI_API_KEY": "sk-..."}
)

print(f"Deployment ID: {result['id']}")
print(f"Status: {result['status']}")
Returns: Dictionary with deployment metadata including:
  • id: Unique deployment identifier
  • status: Deployment status (“BUILDING”, “READY”, “FAILED”)
  • url: Endpoint URL (when status is “READY”)

​
List deployments

Retrieves all deployments associated with the authenticated account. 
deployments = client.deployment.list()

for dep in deployments:
    print(f"{dep['name']}: {dep['status']}")