Verdicts: Ready, Check, Likely broken
Every plugin and theme gets exactly one status. It is worked out from the findings’ severity and confidence:
| Verdict | Meaning | Rule |
|---|---|---|
| Likely broken | Something in this plugin or theme will almost certainly stop working in the iframed editor or on 7.1 list tables. Test it before updating, and check for a newer version. | At least one finding with high severity and high confidence |
| Check | There are findings worth a manual test, but the scanner is less sure or the impact is smaller. | Any other high or medium severity finding |
| Ready | Nothing above low severity was found. | Only low or info findings, or none at all |
The summary table sorts Likely broken first, then Check, then Ready, and within each group alphabetically.
Severity and confidence#
Each finding carries two levels.
- Severity (High, Medium, Low, Info) says how much it matters if the finding is real. The summary table counts findings per severity.
- Confidence (High, Medium, Low) says how sure the scanner is that the match is real. High means the file is known to load in the editor and the pattern is unambiguous. Confidence is lowered by one step when an editor script was identified by a static guess rather than by the runtime check; the finding then shows Lowered: found as an editor script by a static guess.
Upcoming changes#
Some rules cover changes that are not in a WordPress release yet: React 19 and real-time collaboration. Their findings are counted in the Upcoming column and do not affect the verdict by default, so that a plugin is not marked Likely broken for something that cannot break today.
To see how your site would fare once those changes ship, tick Count upcoming changes (React 19, real-time collaboration) in the status above the summary table. The same switch applies to exports made from that view, and the --upcoming flag does the same in WP-CLI.
Newer version available?#
The summary shows Update to x.y.z available next to any plugin or theme for which WordPress already knows a newer version (from its normal update check; the scanner makes no remote request of its own). A newer version may already fix the findings: update it on a staging copy, scan again, then report what is left to the author.