#Upgrading beyond 3.28 causes agents to just ghost me mid response
1 messages · Page 1 of 1 (latest)
This sounds like a real provider/runtime regression, not something I’d try to “fix” by guessing config keys.
First thing I’d do: don’t upgrade to an intermediate build — upgrade straight to current, then repair/migrate:
openclaw --version
openclaw update
openclaw doctor --fix
openclaw gateway restart
openclaw status --all
openclaw models status
Then reproduce with one clean prompt in a new session (/new) and grab logs:
openclaw logs --follow
If it dies mid-response again, please paste:
openclaw --versionbefore + after upgrade- the exact model refs from
openclaw models status//model status - 30–60 log lines around the failure
- whether this happens in a brand-new
/newsession or only old sessions
Given you’re seeing this across Copilot, DeepSeek, and OpenRouter, the interesting bits are likely model transport / streaming / replay shape, not the individual provider key.
Docs for the triage ladder: https://docs.openclaw.ai/help/troubleshooting
Models/auth debugging: https://docs.openclaw.ai/help/faq-models
openclaw --version
OpenClaw 2026.3.28 (f9b1079)
npm i -g openclaw@latest
openclaw --version
OpenClaw 2026.4.25 (aa36ee6)
Now my agents do not respond to any message on telegram
Yep — first split this into Telegram transport/pairing vs agent/model runtime. After a jump from 2026.3.28 → 2026.4.25, also make sure the running gateway/service actually got restarted/reinstalled, not just the CLI.
Run this, then send a Telegram DM while logs are open:
openclaw status --all
openclaw gateway status --deep
openclaw doctor --fix
openclaw gateway install --force
openclaw gateway restart
openclaw channels status --probe
openclaw pairing list telegram
openclaw logs --follow
Things to look for: