pure hound
Hey, we’re seeing recurring OpenClaw run failures:
"The run was interrupted before a response was generated. Please try again."
Logs show:
rawErrorPreview="terminated"
rawErrorHash="sha256:e8c95a2a7271"
provider="openai-codex"
model="gpt-5.5"
modelApi="openai-codex-responses"
Environment:
OpenClaw 2026.4.22
Node v22.22.1
Linux 5.15.0-174-generic
embeddedHarness runtime=auto, fallback=pi
This happens repeatedly across multiple agents/chats, including new Telegram chats/threads, not just old large sessions.
What usually causes rawErrorPreview="terminated" here, and what should we check or log next?