Add reproducible 90-sample security benchmark#2
Merged
Conversation
…cal results Review follow-ups (blocking): - run.mjs imports @flagrix/scanner-core from node_modules instead of the ../../flagrix-scanner-core sibling checkout — the benchmark now runs from a bare clone of this repo, and the devDependency pin actually governs the engine under test - metadata records the installed package version (0.3.0) and the snapshot rules version; sibling git sha/dirty fields are gone, so results can no longer be produced from an unpublished dirty working tree without saying so - post-fix-final regenerated against published scanner-core 0.3.0, rules snapshot 2026.07.12.003, CLI 0.1.4: 90/90 passed, all four gates green, both severe reference findings covered by reference-review.json
- extract shared sample constructors into benchmark/samples.mjs — run.mjs and pin-corpus.mjs previously duplicated projectSample/osvSample and corpus expansion, so the pinner and runner could drift apart silently (verified: expanded corpus reproduces the committed lock exactly, 90/90 sampleIds and fixture contents) - gitignore only ad-hoc results; canonical snapshots (pre-fix-baseline.*, post-fix-final.*) are exempted so updating them never needs git add -f - README: state that the known-malicious-metadata gate measures signature coverage of the pinned OSV samples, not generalizing recall; behavioral generalization is the critical fixtures' job - comment the prefetch's mirrored scanner-core selection constants: drift is correctness-safe (REST fallback), re-sync on dependency bumps
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What changed
@flagrix/scanner-core@^0.3.0Why
Flagrix needed reproducible evidence for malicious recall and false-positive launch gates. The runner never clones, installs, or executes benchmark samples.
Results
Release readiness
Validation