openclaw - 💡(How to fix) Fix Discord delivery silently drops messages after response termination + retry [1 participants]

When an LLM response is terminated mid-execution (e.g., due to timeout or interruption) and automatically retried, the retry may complete successfully (tool calls execute, text response generated) but the final text message fails to deliver to Discord silently — no error log, no retry, user never receives it. 2. If the first LLM response gets terminated (visible in session transcript as stopReason: "error", errorMessage: "terminated") {"stopReason": "error", "errorMessage": "terminated", "timestamp": "2026-03-25T10:06:36.394Z"}

• No delivery error logged in gateway.log for this specific message

1. Gateway should log an error when Discord message delivery fails

Fix Plan

To address the issue of silent failures in Discord message delivery after a successful retry, we need to implement the following steps:

• Modify the Discord gateway to log errors when message delivery fails
• Implement a retry mechanism for failed deliveries
• Notify the user through an alternative path if delivery permanently fails

Code Changes

Here's an example of how you can modify the Discord gateway to log errors and implement a retry mechanism:

import logging
import discord

# Set up logging
logging.basicConfig(level=logging.ERROR)

# Define a function to send a message to Discord
def send_message(message):
    try:
        # Send the message
        discord_channel.send(message)
    except discord.HTTPException as e:
        # Log the error and retry
        logging.error(f"Failed to send message: {e}")
        # Retry the message delivery
        retry_message_delivery(message)

# Define a function to retry message delivery
def retry_message_delivery(message):
    max_retries = 3
    retry_count = 0
    while retry_count < max_retries:
        try:
            # Send the message
            discord_channel.send(message)
            break
        except discord.HTTPException as e:
            # Log the error and increment the retry count
            logging.error(f"Failed to send message (retry {retry_count+1}/{max_retries}): {e}")
            retry_count += 1
    if retry_count == max_retries:
        # Notify the user through an alternative path if delivery permanently fails
        notify_user_alternatively(message)

# Define a function to notify the user through an alternative path
def notify_user_alternatively(message):
    # Implement an alternative notification method, such as sending an email or a notification through another channel
    pass

Verification

To verify that the fix worked, you can test the following scenarios:

• Send a message that triggers a retry and verify that the message is delivered to Discord
• Simulate a failed delivery by disconnecting the Discord WebSocket and verify that the error is logged and the message is retried
• Verify that the user is notified through an alternative path if delivery permanently fails

Extra Tips

• Make sure to handle exceptions properly and log errors to ensure that issues are detected and addressed
• Implement a queueing mechanism to handle messages that fail delivery permanently
• Consider implementing a circuit breaker pattern to detect and prevent cascading failures in the Discord gateway