Repository navigation
docs: don't state the post-merge gate state as the shipped state - #105
Merged
Merged
Conversation
Both READMEs said dismiss/close is held back by 'two locks'. That is true only after the operator merges the permission pack; on a fresh install there is one. The 🛑 box thirty lines below already said so, so each page contradicted itself and the claim stated as fact up front is the one a skimmer keeps. This direction of error is the dangerous one, and the tooling already knows it: install.sh refuses to report a gate ON it cannot verify, and #73 added an exit path for a merge that claims success while gate_on() reads OFF. The front page should not assert what the installer declines to assume. Also corrects the Evidence row: CONTRIBUTING.md names two release gates, both in security/; evals/ is a regression harness, not a gate. Caught in review by Matt Willems on #97. Signed-off-by: Steve Wilson <steve.wilson@exabeam.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Addresses both findings from @mattwillems-exabeam's review on #97.
1. The gate overclaim ⛔ (his blocking finding)
Both READMEs asserted dismiss/close is held back by two locks. True after the permission pack is merged; on a fresh install there is one — the skill's in-prompt ask. The 🛑 box thirty lines below said exactly that, so each page contradicted itself, and the claim stated as fact up front is the one a skimmer retains. The same overclaim appeared twice in
plugin/README.md.Fixed with his suggested shape — qualify the sentence rather than rewrite it:
His argument for holding the page over one clause is the right one and worth preserving: a false "gate is ON" is the dangerous direction, and this repo's tooling already encodes that.
install.shrefuses to report a gate ON it cannot verify and treats "cannot verify" as a distinct third outcome; #73 added an exit path for a merge that reports success whilegate_on()still reads OFF. The front page should not assert what the installer deliberately declines to assume.It also undercut the hero line — "You keep the verdict" is precisely what is not yet true on a fresh install.
2. Evidence row overclaim⚠️
CONTRIBUTING.md:121-124names exactly two release gates — red team and Agent Behavior Verification — both insecurity/.evals/is a regression harness;:214asks only for a recorded run when a fixture or the harness changes. The row now attributes gating to the two gates that exist and describesevals/as what it is.Scope
README.md,plugin/README.md,CHANGELOG.md. 269 invariant tests pass.plugin/README.mdis in the shipped payload, but this is docs-only — no version bump, consistent with Promote new README to main — content review wanted #97's resolved version question.