I would like to see the ability to remove the first-party orgs from the trustedActionsOwnerDataModel, enabling unpinned warnings on first-party actions. If we're mandating SHA pinning on all external actions but GitHub's actions get overlooked, it leaves a gap in our code scanning.
Current usage:
extensions:
- addsTo:
pack: codeql/actions-all
extensible: trustedActionsOwnerDataModel
data:
- ["org1"]
- ["org2"]
Potential (psuedocode, not real working examples):
extensions:
- addsTo:
pack: codeql/actions-all
extensible: trustedActionsOwnerDataModel
data:
- ["org1"]
- ["org2"]
- ["!github"] # this syntax
- ["!actions"]
- ["!advanced-security"]
- removesFrom: # or this syntax?
pack: codeql/actions-all
extensible: trustedActionsOwnerDataModel
data:
- ["github"]
- ["actions"]
- ["advanced-security"]
- addsTo: # or this?
pack: codeql/actions-all
extensible: distrustedActionsOwnerOverrideDataModel # new
data:
- ["github"]
- ["actions"]
- ["advanced-security"]
I would like to see the ability to remove the first-party orgs from the trustedActionsOwnerDataModel, enabling unpinned warnings on first-party actions. If we're mandating SHA pinning on all external actions but GitHub's actions get overlooked, it leaves a gap in our code scanning.
Current usage:
Potential (psuedocode, not real working examples):