#⚠️ Something went wrong while processing your request. Please try again, or use /new to start a fres
9 messages · Page 1 of 1 (latest)
Let me check the docs and search for that exact failure string so I can give you something concrete instead of guessing.
I found the relevant code path too: that exact banner is OpenClaw’s generic fallback when the upstream model run fails and it can’t classify the error more specifically. So we should treat it as “check gateway/model logs”, not as a Discord-only symptom.
That banner is annoyingly generic. It means the model run failed, not necessarily that Discord is broken.
Because OpenClaw is able to send the warning back into Discord, the Discord API side is probably fine. The failure is usually one of these:
- stale/corrupted session state
- Anthropic auth or quota issue
- upstream model error
Try this first:
-
Send
/newas a standalone message in that chat.- If that fixes it, the session got out of sync.
-
If it still fails, paste these outputs:
openclaw --version
openclaw status --all
openclaw gateway probe
openclaw gateway status
openclaw doctor
openclaw channels status --probe
openclaw models status
- Then do:
openclaw logs --follow
Send one test message from Discord, then paste the relevant error lines.
Optional but very useful isolator:
openclaw infer model run --model anthropic/claude-sonnet-4-6 --prompt "hi" --json
If that fails too, it’s not Discord, it’s the model/auth path.
Common log clues to watch for:
rate_limit_errororExtra usage is required for long context requestsNo API key found for provider "anthropic"Model login expired on the gatewaymessages.content.tool_use.input field requiredContext overflow
Docs: