What it does
Cached results let you change the confidence bar without asking again. Its experimental code-suppression judgments have not been audited, and its auto-merge score is only displayed: it never opens or merges a fix.
Agent tooling
vulnwash sorts Snyk security findings into review buckets, combining fixed dependency rules with Jev’s judgments about risk and code.
Only you can see your notes.
Screenshot unavailable. Open the experiment ↗
Cached results let you change the confidence bar without asking again. Its experimental code-suppression judgments have not been audited, and its auto-merge score is only displayed: it never opens or merges a fix.
If your project has a long security report, ask your developer to try a saved report first. Compare the tool’s groups with the original warnings, including those it sets aside. Changing how sure it must be uses saved answers; it does not check the code again.
The tool sends details about software your project uses, and nearby source lines for code warnings, to TypeSafe, an outside AI service. Get permission before sending private code. A person should check dismissed warnings: the maker has not checked how often these dismissals are wrong. A reassuring score never merges a fix.