ZacharyZhang-NYandClaude Opus 5.5
c5cb153f72
A document can forge both a console message and its source (KaTeX's \errmessage, data: scripts, sourceURL, encoded paths), so the display handler logs a line number and nothing the page wrote. The page's own diagnostics reach the app over elyQuery: a missed reveal sends its line; the visual failure and diagram lines already did. A test calls the real handler and checks the captured log. Round 5 found only the missing test; it is added and mutation-checked, with no sixth review under the five-round cap. M-15 is new: the print page runs a document's scripts. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1.0 KiB
1.0 KiB
FAIL
- crates/ely-web-content/src/handlers.rs:273 — The new console privacy boundary has no repository regression test; the previous
Surface::scripttest was removed without replacement. In a review copy, restoring logging of_messageand_sourceleaves all 32 ElyMark/Web-content Cargo tests passing on WSL, although a warning carrying document text now writes it verbatim to logs, violating PRD §18.2 and rule 4. A probe capturing this actual callback's log records passes with the current implementation and fails with that mutation on both Windows and WSL. Retain such a regression in the repository: private markers in warning/error messages and arbitrary sources must never reach logs, while the line diagnostic remains observable. Evidence:.work/m11-r5/linux-mutant-repository.log,windows-mutant-probe.log,linux-mutant-probe.log, andevidence.log. Both round 4 leaks are fixed, confirmed by native CEF probes on both platforms. Probe copies and their dedicated targets were deleted before this verdict.