Repository navigation
Five rules have two requirement levels but one force value #193
Replies: 1 comment
|
These edge cases in particular have always been a bit of a complicated thorn - they are specific situations where we felt that having the secondary requirement directly inside the rule was critical because it applied specifically to the rule in a way that it needed to be directly in there to avoid confusion. The trick is that they're secondary, contingent on the first part, so the core force of the rule is still the main thing that matters for determining the approach to implementation.
So I think the right approach for CR26 maintenance is your third suggestion "Add a note to AGENTS.md that force is the primary level but the statement can carry a second one (doesn't touch the data)." And to explain this in Using the FedRAMP Consolidated Rules. Created FedRAMP/rules#54 to track this, thank you! |
Uh oh!
There was an error while loading. Please reload this page.
Five rules in 2026.11.02.01 use two requirement keywords in one statement, and
forceinfedramp-consolidated-rules.jsoncan only have one.forceAnything that reads
forceto build a requirement list gets half of each rule. For example, filtering on MUST misses the MUST in MAS-CSO-SUP and SCN-CSO-EMG, because those are filed under MAY.The new rule proposed in FedRAMP/rules#52 (IEC-CSO-RID) looks the same "MUST NOT ... but MUST" pattern as VER-RPT-NID, so that would be another.
I'm not sure what the correct way to handle it should be, but I can think of a few options:
forcehave more than one value.forceis the primary level but the statement can carry a second one (doesn't touch the data).I'm mentioning this because #190 asks for a flat list of rule IDs with the force for each, and that list would miss the second keyword for these rules.
Thanks for your great work!
All reactions