hit 429 Too Many Requests on the third tool call in a four-node graph. Retry-After said 18s. cool.
the checkpointer still wrote the interrupted state. next resume picked up with messages missing the last tool result, so the agent just... invented one. looked plausible. was wrong.
anyone gating resumes on a full tool-call roundtrip, or am i supposed to delete the checkpoint by hand every time groq hiccups?
left a Roo Code session running Friday night against a LangSmith eval suite. one case was flaky — TimeoutError: run exceeded 60s — and the agent just… kept retrying it. no max attempts. no spend cap.…
turned on Auto Labels last week because my inbox was a landfill. this morning a customer pinged me on Slack about a chargeback. i searched Stripe. nothing in Primary. found the dispute email sitting…
Told Cursor agent to "clean up unused scripts". It removed .github/workflows/deploy.yml and the agent deny comment at the top of scripts/break glass.sh . PR looked fine until staging had no deploy jo…
2 comments
Join the discussion
Log in to comment.
yeah that invented tool result is the scary part. green resume, wrong side effects.
i started rejecting resume if any
tool_callsentry has no matchingtoolmessage. cheap check. saved me once when Stripe got a phantom refund arg from a half node.i delete the checkpoint. every time.
feels dumb until you watch an agent "finish" a payment step that never actually ran. rate limits and checkpointers do not mix without a human kill switch.