Daita Client
A lightweight Python SDK for triggering and monitoring your deployed Daita agents from any project.
#What is Daita Client?
daita-client is a standalone Python package that lets you run your deployed agents and workflows programmatically — from a web app, a script, a notebook, or any other Python project. It handles authentication, execution, and result polling, so you can integrate your hosted agents in minutes.
You don't need the full Daita agent framework installed to use it.
#Installation
pip install daita-clientIf you already have daita-agents installed, daita-client is included and importable from the same place.
#Authentication
Initialize the client with your API key from the Daita dashboard. You can also set DAITA_API_KEY as an environment variable and omit it from the constructor entirely.
from daita_client import DaitaClient
client = DaitaClient(api_key="your_api_key")#Run Your First Agent
from daita_client import DaitaClient
client = DaitaClient(api_key="your_api_key")
# Execute an agent and wait for the result
result = client.execute_agent("my_agent", data={"query": "summarize Q1 sales"}, wait=True)
print(result.status) # "completed"
print(result.result) # {"answer": "..."}The wait=True parameter blocks until the execution finishes. Without it, the call returns immediately with an execution_id you can use to poll for the result later.
#Execution Result
Every method returns an ExecutionResult:
| Field | Description |
|---|---|
execution_id | Unique ID for this execution |
status | queued, running, completed, failed, or cancelled |
target_name | Name of the agent or workflow that ran |
result | Output returned by the agent |
error | Error message if the execution failed |
created_at | When the execution was queued |
completed_at | When the execution finished |