Hey, thanks for the report with the Request ID. It helped us find the cause quickly.
The issue is that the context from this repo is over the model token limit 200k in non-max mode. Cursor tries to shrink the context with summarization, but it is not enough, so it keeps retrying until it runs out of retries. That is why it works in other projects where the context is smaller.
Workarounds:
- Start a new chat instead of continuing a long one, long sessions build up context
- Try MAX mode, it supports a higher token limit
- Split tasks into smaller requests so the agent does not pull in too much context at once
- If you are using Composer 2, try switching to a specific model like Claude Sonnet 4.6 or GPT-5
The team is aware of this issue. The problem is that summarization cannot recover after going over the limit. There is no ETA for a fix yet, but your report helps us prioritize it.
Let me know if these workarounds helped.