Skip to main content

​
Error Handling

Chat includes built-in error handling with automatic retries for transient failures.

​
Retry Configuration

from upsonic import Agent, Chat

agent = Agent("openai/gpt-4o")

chat = Chat(
    session_id="session1",
    user_id="user1",
    agent=agent,
    retry_attempts=3,
    retry_delay=1.0
)

​
Error States

Chat tracks session state and handles errors: 
from upsonic import Agent, Chat

agent = Agent("openai/gpt-4o")
chat = Chat(session_id="session1", user_id="user1", agent=agent)

# Check session state
if chat.state.value == "error":
    # Handle error state
    chat.reset_session()

​
Handling Errors

from upsonic import Agent, Chat

agent = Agent("openai/gpt-4o")
chat = Chat(session_id="session1", user_id="user1", agent=agent)

try:
    response = await chat.invoke("Hello")
except Exception as e:
    print(f"Error: {e}")
    # Reset session if needed
    chat.reset_session()

​
Retryable Errors

Chat automatically retries on:
  • Network errors (ConnectionError, TimeoutError)
  • Rate limiting
  • Temporary service unavailability
  • Internal server errors
Non-retryable errors (validation errors, authentication failures) are raised immediately.

​
Concurrent Invocation Limits

from upsonic import Agent, Chat

agent = Agent("openai/gpt-4o")

chat = Chat(
    session_id="session1",
    user_id="user1",
    agent=agent,
    max_concurrent_invocations=2
)

# Can handle up to 2 concurrent invocations
# Additional calls will raise RuntimeError

​
Error Recovery

from upsonic import Agent, Chat

agent = Agent("openai/gpt-4o")
chat = Chat(session_id="session1", user_id="user1", agent=agent)

try:
    response = await chat.invoke("Hello")
except RuntimeError as e:
    if "error state" in str(e):
        chat.reset_session()
        response = await chat.invoke("Hello")