pyOpenSci / buffy

Editorial bots generator used for pyosbot
https://pyos-bot-buffy.readthedocs.io/en/latest/
MIT License
1 stars 1 forks source link

How to figure out what in the config yaml file broke the bot? #7

Closed lwasser closed 7 months ago

lwasser commented 8 months ago

hey @xuanxu i sent you a slack message but it might be nice to chat here openly. it seems like there are specific things in the YAML file that cause the bot to not respond. However i'm not sure how to identify what i've done wrong to break things.

i wondered if you could help me understand how to troubleshoot? it's probably hard to see if all tests pass but the bot doesn't respond. One issue was the steps having the same name. i think i've fixed that. but now it's broken again and i'm not sure what i did. i obviously did something however.

xuanxu commented 8 months ago

Hey! When the bot responds to simple commands (help, hi...) but there is a specific command where it goes silent, then probably that command is wrongly configured. But when the bot does not respond to anything usually there is something wrong in the YAML file. The two more common causes of YAML-parsing errors are:

Currently the test suite covers the internal code, it runs the Buffy' specs as a bots generator, it does not validate individual bot config files.

xuanxu commented 8 months ago

Another tool for debugging is the webhook delivery log, where you can see the server response and redeliver any past request. https://github.com/pyOpenSci/test-pyos-review/settings/hooks

lwasser commented 8 months ago

thank you @xuanxu !!! i understand now!

lwasser commented 7 months ago

closing this as answered!