Call your deployed agents with the Reminix Python SDK. Invoke agents and stream responses in real-time.
Installation
Configuration
Basic Setup
from reminix import Reminix
# Using API key directly
client = Reminix(api_key="reminix_sk_...")
# Using environment variable (recommended)
import os
client = Reminix(api_key=os.environ.get("REMINIX_API_KEY"))
The SDK automatically reads from the REMINIX_API_KEY environment variable if no key is provided.
Configuration Options
from reminix import Reminix
client = Reminix(
api_key="reminix_sk_...",
# Custom base URL (for self-hosted or testing)
base_url="https://api.reminix.com/v1",
# Request timeout in seconds
timeout=60,
# Custom headers
default_headers={"X-Custom-Header": "value"}
)
Quick Example
from reminix import Reminix
client = Reminix()
# Invoke a task-oriented agent
response = client.agents.invoke(
"my-agent",
prompt="Analyze this data"
)
print(response["content"])
# Invoke a chat agent
response = client.agents.invoke(
"chat-assistant",
messages=[{"role": "user", "content": "Hello!"}]
)
print(response["message"]["content"])
Async Support
For async applications, use the AsyncReminix client:
import asyncio
from reminix import AsyncReminix
async def main():
async with AsyncReminix() as client:
# Async invoke
response = await client.agents.invoke(
"my-agent",
prompt="Analyze this"
)
# Async chat
response = await client.agents.invoke(
"chat-assistant",
messages=[{"role": "user", "content": "Hello!"}]
)
asyncio.run(main())
The async client uses async with for proper connection management. You can also instantiate it directly and call await client.close() when done.
Next Steps
