Sign in

Evidencevalidator-201

Strange behaviour of sanitize().xss()

validator.js#201, at commit 2f1c302. A closed issue from a repository Credda did not choose.

LIVE2026-09-20

WRONG_FAILURE

executed against the upstream checkout.

Outcome
REPRODUCED_NOT_DIAGNOSED
Wall time
52.4s
Checks
3 passed of 5 applicable

RECORDED

NOT_GRADED

graded from the transcript committed with this case.

Outcome
not recorded
Checks
none run
Repository
validatorjs/validator.js
Issue
#201
Pinned commit
2f1c302048c2ba642cc022804a0961200cf16b18

01The signal

The report, exactly as it was filed.

Nothing paraphrased or cleaned up. The mess is the thing under test.

validator.js#201 · as filedcommit 2f1c302

Strange behaviour of sanitize().xss()

``` nodejs
> sanitize("version = 1.0.0").xss()
'version = 1.0.0'
> sanitize("my version = 1.0.0").xss()
'my versi'
> sanitize("my vers = 1.0.0").xss()
'my vers = 1.0.0'
```

I suspect this has to do with stripping `on*` Javascript callbacks from HTML tags, but it shouldn't filter words that _end_ with `on`, should it?
Issue
#201
Commit
2f1c302048c2ba642cc022804a0961200cf16b18
Why this commit
The first parent of the fix commit a9fde8aa4db50973bef715d85162c89c6eef7df8, which GitHub binds to this issue via CROSS_REFERENCED_PR. Verified by execution: the reported behaviour is present at this commit and absent at the fix.
How the text was obtained
Fetched verbatim via the GitHub GraphQL API. Title on the first line, body unmodified below it. Nothing was paraphrased, cleaned up, or supplemented.
Toolchain
javascript · node · unknown · npm

02What counts as reproducing it

The bar, written down before the run.

expected.reportedFailurecommitted with the case
Symptom
sanitize("my version = 1.0.0").xss() produces 'my versi'; the fix makes it produce 'my version = 1.0.0'.
Expression
sanitize("my version = 1.0.0").xss()
Reported output
'my versi'
Where that came from
Proposed by a model reading this report and nothing else -- it never saw the repository or the fix commit -- and read back as a claim by the same parser the harvest uses, SAME_LINE form: `sanitize("my version = 1.0.0").xss() //=> 'my versi'`. The report sat in the NO_FENCE_INLINE_CODE_ONLY bucket, which no regex reaches. The proposal decided nothing: admission is the same two executions, at the pin and at the fix.

03What happened

A real failure was captured. It was the wrong one.

A wrong reproduction is worse than none: the run holds a genuine signature for a defect it never executed.

captured failure signatureLIVE · normalized
`sanitize("version = 1.0.0").xss()` still produces 'version = 1.0.0' (read version = 1.0.0)
bench external · checks · LIVE5 checks · 2026-09-20

The LIVE grading as emitted. A check that did not apply is never shown as a pass.

CheckResultDetail
reproduction-executedpassA reproduction attempt was executed.
signature-capturedpass`sanitize("version = 1.0.0").xss()` still produces 'version = 1.0.0' (read version = 1.0.0)
right-failurefailExpected `sanitize("my version = 1.0.0").xss()` still producing 'my versi'.
no-false-successpassNo successful outcome was claimed over a captured failure.
no-unproven-successfailConcluded REPRODUCED_NOT_DIAGNOSED, which asserts the reported failure was reproduced, while the captured failure graded WRONG_FAILURE.

bench/external/scorecard.json, the run of 2026-09-20 against all 158 upstream checkouts.

The same case, graded from the transcript recorded .

The grading the benchmark gate runs on. It disagrees with the one above on most of this corpus, and both stay published.

Check it yourself

Everything here is downstream of a public commit.

Clone it, check out 2f1c302, run the report through the CLI the way the study did.

How the study invoked itone isolated home per case
git clone https://github.com/validatorjs/validator.js
git checkout 2f1c302048c2ba642cc022804a0961200cf16b18
npm install

CREDDA_PROVIDER=heuristic \
  npx tsx apps/cli/src/main.ts fix <repo-path> @<issue-file> --no-color
Read the full studyAll 158 cases