Enable pytorchbot on vLLM

[huydhn]

This is a request from vLLM maintainers (@youkaichao @simon-mo) to enable HUD and Dr.CI on vLLM repo. This would also allow us to save vLLM GitHub events into ClickHouse and make them available for query, i.e. via Flambeau. The task of enabling pytorchbot on vLLM is simple, just need to request @simon-mo to install https://github.com/apps/pytorch-bot on vLLM

1. The bot should be added ONLY to public projects like vllm-project/vllm
2. Some logics on HUD and Dr.CI might not work with a non-pytorch repo and we need to iron them out. The good news is that I expect the fixes are straightforward

One note is that mergebot, as part of pytorchbot, won't work on vLLM, but it's out of scope of this issue because it's not an easy task to roll it out to repos beside pytorch/pytorch. There is no need to enable it on vLLM for now. It's a no-op because there is already a check preventing mergebot from running there

cc @jeanschmidt @seemethere

[simon-mo]

Installed

[huydhn]

Some features are spammy and not relevant to vLLM (yet), it's ciflow in this case vllm-project/vllm#24673. ciflow is needed as part of GHA migration, so it seems like we need to:
(1) Either gate this feature and prevent it from running on vLLM
(2) or enable the bot once GHA migration work finishes ciflow support (cc @jeanschmidt )