[BUG]: Some checks are not being executed on all merges

[claudioandre-br]

Which version of the application are you using?

1.9J1+2404, latest bleeding

Description of the bug

Some checks are not being executed on all merges. We have already detected problems in these cases:

• Trivy rootfs
• Trivy SBOM
• Supply chain / scorecard

Sometimes I'm pushing to main and sometimes not?

GitHub isn't doing what it should. Anyway, we can think of an alternative solution (workaround).

Examples:

• 15 checks
  6579932

• 11 checks
  cdb3a83

The difference between these two points in time:

https://github.com/openwall/john-packages/compare/6579932baa008b74512119bbaa250427c6d1313c..cdb3a83e872a3ceb34fb9216d12c8372ca45bc26

Nothing that could be our fault.

Steps To Reproduce

No response

Your execution environment

No response

[github-actions]

Welcome and thank you for helping us improve our software.

Please make sure you read and understand the CODE_OF_CONDUCT.md and CONTRIBUTING.md.
We hope to see you soon!

[claudioandre-br]

More examples. History for scorecard.yml:

• Scorecard has been executed:
  df5f3df

• Scorecard has NOT been executed:
  abee9d7

• Diff:
  https://github.com/openwall/john-packages/compare/df5f3df3a72dd95ca7adb7bb0510d3b92782d1d8..abee9d73ee3fbdb32b8bc9e9980538e9fa84da68

[claudioandre-br]

More examples:

• 13 checks (4 skipped checks):
  ba808f3

• 12 checks (4 skipped checks):
  55d626c

The event that did not occur (and it has not changed at all):

john-packages/.github/workflows/greetings.yml

Lines 26 to 30 in 55d626c

	"on":
	pull_request_target:
	types: [opened]
	issues:
	types: [opened]

1. So a “pull request event” did not happen in the second run?
2. Is there a cache (where) ?
3. If a file has a name as a greetings, does some magic happen behind the curtains? 🪄 😱

[claudioandre-br]

I will run yaml linters on all workflow files (and I'll accept the suggested corrections).

[claudioandre-br]

• It seems that if I use a branch name like fix/name, I can get around the problem. => FALSE [1]
• The merges made by the merge bot are worse. => FALSE [2]

[1]
It might help a little on some occasions.

[2]
The examples above prove this.

[claudioandre-br]

One commit was skipped:

The second commit below is simply IGNORED (the 646):

[claudioandre-br]

I've documented the behavior.
There is nothing I can do to improve it or resolve the situation.